Try using this program to convert the data.
 
http://lcni.uoregon.edu/~jolinda/MRIConvert/
 

Choose the folder with the dicom files in the "Add Folder" as the Input folder (or by clicking "Input > Add folder").


Choose Output type as "FSL Nifti" and in the "Options" button, choose "save multivolume series as 4D files" and "save as .nii file"

It will take a upto 5 minutes to convert.

--
Regards,
Mithra

Imaging Research Center
University of Texas, Austin
3925-B W Braker Lane
Austin. TX - 78759
Tel: (512) 232-5725
Fax: (512) 232-4202

www.irc.utexas.edu


 
On 10/7/07, Deepak Madhavan <[log in to unmask] > wrote:
Hello all:

I am a new FDT user, and I have been having trouble with the eddy correction
function. I used MRIcron to convert my GE 3T DTI dicoms into both nifti and
analyze formats, but when I attempt to run eddy correction, it returns the
error message, 'Input does not exist or is not in a supported format'.  ANy
assistance would be greatly appreciated.

Deepak



------=_Part_40102_28639463.1191786181280-- ========================================================================= Date: Sun, 7 Oct 2007 17:24:12 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Paymon <[log in to unmask]> Subject: Feat error Comments: cc: Mariel Kozberg <[log in to unmask]>, Olivier Piguet <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Hi when i run Feat, i get the following error message: bad window path name ".dialog1.f.viewport.f.btn" bad window path name ".dialog1.f.viewport.f.btn" while executing "winfo rootx $win" (procedure "balloonhelp_show" line 7) invoked from within "balloonhelp_show .dialog1.f.viewport.f.btn" ("after" script) the error message pops up after i click on the design figure that pops up after setting up a model under the "stats" tab. i am running FSL 4.0.1 on a 64bit CentOS 4.4 machine. thank you in advance for any advice. best, paymon ========================================================================= Date: Mon, 8 Oct 2007 11:45:12 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Lea Krugel <[log in to unmask]> Subject: applyxfm4d versus applyxfm In-Reply-To: <[log in to unmask]> Mime-version: 1.0 Content-type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit Dear FSL people, I'd like to know whether "applyxfm4D" with only one transformation matrix for all the volumes of a functional data set leads to somehow different results than flirt in combination with "applyxfm" and "init". I'm asking because I used "applyxfm" for my functional data and FEAT has no problems in recognising the correct number of time points, so I guess the "applyxfm" command works fine here, also for functional data. My understanding of "applyxfm4D" would then be that it is the right tool when applying more than one transformation matrix to the data, but with only one there is no difference between "applyxfm" and "applyxfm4D". Is that right or might there be problems with my output in later analysis steps? Thanks a lot, Lea ------------------------------------------- Lea Krugel Max Planck Institute for Human Development Neurocognition of decision making Lentzeallee 94 14195 Berlin Tel.: +49-30-82406618 Fax: +49-30-82406616 email: [log in to unmask] ========================================================================= Date: Mon, 8 Oct 2007 11:28:42 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Brandon Whitcher <[log in to unmask]> Subject: Constrain FLIRT in sch3Dtrans_3dof Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Dear FSL experts, I am performing a 3D-to-3D registration in FLIRT using the schedule for translations only (sch3Dtrans_3dof).=20=20 Question =3D How can I limit the amount of possible translations in a giv= en application of FLIRT? Some of my registrations are producing crazy translations. thank-you, Brandon ========================================================================= Date: Mon, 8 Oct 2007 12:31:35 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Agnieszka Burzynska <[log in to unmask]> Subject: TBSS on non-FA In-Reply-To: <[log in to unmask]> Mime-version: 1.0 Content-type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit Dear all, I had a problem running randomise for the second set of non-FA data: The error looked similar to: Error:Loading data: termiante called after hortwing an instance of `RBD_COMMON Base Exception' Abort Only when I created a new folder without any folders from tbss_non_FA script, then run the tbss_non_FA script for this set of data, did the randomise work without complains. Did any of you encounter a similar problem? Thanks a lot, Aga ========================================================================= Date: Mon, 8 Oct 2007 11:45:14 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Marc Dubin <[log in to unmask]> Subject: Running tbss_2_reg on 2 processors Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain Hello FSL Users, I have been trying to run a 2nd "tbss_2_reg -n" in a 2nd terminal on my d= ual core mac to speed=20 things up in the NxN registration. The first "tbss_2_reg -n" is running f= ine. The second one, which I=20 run from the TBSS working directory as well, quickly crashes with:=20 marc-dubins-computer:~/Data/sumit31 marc$ tbss_2_reg -n wc: .commands: open: No such file or directory 13866 Any suggestions would be greatly appreciated! Best, Marc ========================================================================= Date: Mon, 8 Oct 2007 13:18:35 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Juergen Haenggi <[log in to unmask]> Subject: Re: Running tbss_2_reg on 2 processors In-Reply-To: <[log in to unmask]> Mime-version: 1.0 Content-type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit Hi Marc We had the same problem. Dual core does not work. You must have two real processors. Only one processor with 2 cores did not work. Best regards Juergen On 8.10.2007 12:45 Uhr, "Marc Dubin" <[log in to unmask]> wrote: > Hello FSL Users, > > I have been trying to run a 2nd "tbss_2_reg -n" in a 2nd terminal on my dual > core mac to speed > things up in the NxN registration. The first "tbss_2_reg -n" is running fine. > The second one, which I > run from the TBSS working directory as well, quickly crashes with: > > marc-dubins-computer:~/Data/sumit31 marc$ tbss_2_reg -n > wc: .commands: open: No such file or directory > 13866 > > Any suggestions would be greatly appreciated! > > Best, > Marc ----------------------------------------------------------- Juergen Haenggi Ph.D. (Dr. des.) Division of Neuropsychology Institute of Psychology University of Zurich Binzmuehlestrasse 14, PO Box 25 8050 Zurich, Switzerland 0041 44 635 73 97 (phone office) 0041 76 445 86 84 (phone mobile) 0041 44 635 74 09 (fax office) BIN 4.D.04 (office room number) [log in to unmask] (email) http://www.psychologie.uzh.ch/neuropsy/ (website) http://www.juergenhaenggi.ch (private website) ----------------------------------------------------------- ========================================================================= Date: Mon, 8 Oct 2007 12:40:03 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: Running tbss_2_reg on 2 processors In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - no, that's not right I'm afraid. Please see: http://www.jiscmail.ac.uk/cgi-bin/webadmin? A2=ind0708&L=FSL&P=R53996&I=-3&X=31C16808CF4918D3CB&Y=steve% 40fmrib.ox.ac.uk You can't run this script multiple times in parallel in FSL4.0; to get parallelisation you need to use SGE. Cheers. On 8 Oct 2007, at 12:18, Juergen Haenggi wrote: > Hi Marc > > We had the same problem. Dual core does not work. You must have two > real > processors. Only one processor with 2 cores did not work. > > Best regards > Juergen > > > On 8.10.2007 12:45 Uhr, "Marc Dubin" <[log in to unmask]> wrote: > >> Hello FSL Users, >> >> I have been trying to run a 2nd "tbss_2_reg -n" in a 2nd terminal >> on my dual >> core mac to speed >> things up in the NxN registration. The first "tbss_2_reg -n" is >> running fine. >> The second one, which I >> run from the TBSS working directory as well, quickly crashes with: >> >> marc-dubins-computer:~/Data/sumit31 marc$ tbss_2_reg -n >> wc: .commands: open: No such file or directory >> 13866 >> >> Any suggestions would be greatly appreciated! >> >> Best, >> Marc > > > ----------------------------------------------------------- > Juergen Haenggi > Ph.D. (Dr. des.) > Division of Neuropsychology > Institute of Psychology > University of Zurich > Binzmuehlestrasse 14, PO Box 25 > 8050 Zurich, Switzerland > 0041 44 635 73 97 (phone office) > 0041 76 445 86 84 (phone mobile) > 0041 44 635 74 09 (fax office) > BIN 4.D.04 (office room number) > [log in to unmask] (email) > http://www.psychologie.uzh.ch/neuropsy/ (website) > http://www.juergenhaenggi.ch (private website) > ----------------------------------------------------------- ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Mon, 8 Oct 2007 14:02:34 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Juergen Haenggi <[log in to unmask]> Subject: Re: Running tbss_2_reg on 2 processors In-Reply-To: <[log in to unmask]> Mime-version: 1.0 Content-type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit Hi Steve Sorry for my wrong information. I conclude this from what we have observed on your computers. Sorry Best regards Juergen On 8.10.2007 13:40 Uhr, "Steve Smith" <[log in to unmask]> wrote: > Hi - no, that's not right I'm afraid. Please see: > http://www.jiscmail.ac.uk/cgi-bin/webadmin? > A2=ind0708&L=FSL&P=R53996&I=-3&X=31C16808CF4918D3CB&Y=steve% > 40fmrib.ox.ac.uk > You can't run this script multiple times in parallel in FSL4.0; to > get parallelisation you need to use SGE. > Cheers. > > On 8 Oct 2007, at 12:18, Juergen Haenggi wrote: > >> Hi Marc >> >> We had the same problem. Dual core does not work. You must have two >> real >> processors. Only one processor with 2 cores did not work. >> >> Best regards >> Juergen >> >> >> On 8.10.2007 12:45 Uhr, "Marc Dubin" <[log in to unmask]> wrote: >> >>> Hello FSL Users, >>> >>> I have been trying to run a 2nd "tbss_2_reg -n" in a 2nd terminal >>> on my dual >>> core mac to speed >>> things up in the NxN registration. The first "tbss_2_reg -n" is >>> running fine. >>> The second one, which I >>> run from the TBSS working directory as well, quickly crashes with: >>> >>> marc-dubins-computer:~/Data/sumit31 marc$ tbss_2_reg -n >>> wc: .commands: open: No such file or directory >>> 13866 >>> >>> Any suggestions would be greatly appreciated! >>> >>> Best, >>> Marc >> >> >> ----------------------------------------------------------- >> Juergen Haenggi >> Ph.D. (Dr. des.) >> Division of Neuropsychology >> Institute of Psychology >> University of Zurich >> Binzmuehlestrasse 14, PO Box 25 >> 8050 Zurich, Switzerland >> 0041 44 635 73 97 (phone office) >> 0041 76 445 86 84 (phone mobile) >> 0041 44 635 74 09 (fax office) >> BIN 4.D.04 (office room number) >> [log in to unmask] (email) >> http://www.psychologie.uzh.ch/neuropsy/ (website) >> http://www.juergenhaenggi.ch (private website) >> ----------------------------------------------------------- > > > ------------------------------------------------------------------------ > --- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > ------------------------------------------------------------------------ > --- ----------------------------------------------------------- Juergen Haenggi Ph.D. (Dr. des.) Division of Neuropsychology Institute of Psychology University of Zurich Binzmuehlestrasse 14, PO Box 25 8050 Zurich, Switzerland 0041 44 635 73 97 (phone office) 0041 76 445 86 84 (phone mobile) 0041 44 635 74 09 (fax office) BIN 4.D.04 (office room number) [log in to unmask] (email) http://www.psychologie.uzh.ch/neuropsy/ (website) http://www.juergenhaenggi.ch (private website) ----------------------------------------------------------- ========================================================================= Date: Mon, 8 Oct 2007 13:57:41 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Dave Flitney <[log in to unmask]> Subject: Re: fslview: segmentation fault In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: multipart/alternative; boundary=Apple-Mail-3-940180267 --Apple-Mail-3-940180267 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Naj, It opens fine on the platforms I've tried so far - FSLView 3.0 on: MacOSX (PPC and Intel); Fedora Core 5 (64bit); and Centos4 (32bit). Which OS & FSLView versions are you using? On 5 Oct 2007, at 16:51, Najmeh Khalili M. wrote: > Hi Dave, > > Ref number is: > 943835 > > This is supposed to be a mask. It is not empty at fslstats > indicates non-zero content. Didn't think it was. That was a different poster :-) > Thanks > Naj -- Cheers, Dave Dave Flitney, IT Manager Oxford Centre for Functional MRI of the Brain E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 URL: http://www.fmrib.ox.ac.uk/~flitney --Apple-Mail-3-940180267 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1 Naj,

It opens fine on the = platforms I've tried so far - FSLView 3.0 on: MacOSX (PPC and Intel); = Fedora Core 5 (64bit); and Centos4 (32bit).=A0

Which OS & FSLView = versions are you using?

On 5 Oct 2007, at 16:51, = Najmeh Khalili M. wrote:

Hi Dave,

Ref number is:
943835

This is supposed to be a mask. = It is not empty at fslstats
indicates = non-zero content.

Didn't think it was. That was a = different poster :-)

Thanks
Dave Flitney, = IT Manager
Oxford Centre for Functional MRI = of the Brain
E:[log in to unmask] = W:+44-1865-222713 F:+44-1865-222717

=

= --Apple-Mail-3-940180267-- ========================================================================= Date: Mon, 8 Oct 2007 14:22:56 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Dave Flitney <[log in to unmask]> Subject: Re: Problem with GUI? In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: multipart/alternative; boundary=Apple-Mail-4-941695492 --Apple-Mail-4-941695492 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Just had a re-read of the docs. The path aliasing feature appears to be used to resolve problems with current working directory only, i.e., not general path aliasing (See page 102 of ng1e6-admin.pdf). If this is the case, then you can't use this feature to re-map paths in command lines. You'll have to provide all the required paths or re- factor as we did to make all mount points the same across our cluster nodes and submission hosts. On 5 Oct 2007, at 20:16, Lokke Highstein wrote: > On Oct 5, 2007, at 5:45 AM, Dave Flitney wrote: > >> Hi Lokke, >> >> On 4 Oct 2007, at 22:11, Lokke Highstein wrote: >> >>> one thing is that the paths to the data we are working on are >>> slightly different on the head nodes and the cluster nodes, due >>> to the head node having the data mounted off of an xraid (and >>> showing up in /Volumes, which i then made symbolic links to the / >>> directory) and the nodes mounting the same directories at /nfs/ >>> with symbolic links to / >> >> Could you confirm: >> >> On the head node /data_directory is a link to /Volumes/data_directory >> On compute node /data_directory is a link to /nfs/data_directory > > well i was using placeholder names (data_directory) but the actual > data in this case lives in /Volumes/Clinical/data/lokke_test/ > tky_13dir on the head node. > > also on the head node /data is a link to /Volumes/Clinical/data > > on the cluster node it's in /nfs/Clinical/data/lokke_test/tky_13dir > > and on that node /data is a link to /nfs/Clinical/data > > FYI the use of the name "Clinical" is just bad naming, it isn't > actually clinical data there, but not it's hard for me to change > the name at this point (long story, not worth telling.) > >>> i have set up sge_aliases which are supposed to solve this, but >>> still it seems that when we submit a bedpost job through the GUI >>> (which forces the full path to be used - even if we type in the / >>> data_directory path it then resolves the actual full path) it >>> ends up in an error state in the queue. >> >> Could you send me the output of: 'cat ${SGE_ROOT}/default/common/ >> sge_aliases' >> Also an example of the full error message (try 'qstat -j - >> xml | grep -i QIM_message') > > > #___INFO__MARK_BEGIN__ > ###################################################################### > #### > # > # The Contents of this file are made available subject to the > terms of > # the Sun Industry Standards Source License Version 1.2 > # > # Sun Microsystems Inc., March, 2001 > # > # > # Sun Industry Standards Source License Version 1.2 > # ================================================= > # The contents of this file are subject to the Sun Industry Standards > # Source License Version 1.2 (the "License"); You may not use this > file > # except in compliance with the License. You may obtain a copy of the > # License at http://gridengine.sunsource.net/ > Gridengine_SISSL_license.html > # > # Software provided under this License is provided on an "AS IS" > basis, > # WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, > INCLUDING, > # WITHOUT LIMITATION, WARRANTIES THAT THE SOFTWARE IS FREE OF > DEFECTS, > # MERCHANTABLE, FIT FOR A PARTICULAR PURPOSE, OR NON-INFRINGING. > # See the License for the specific provisions governing your > rights and > # obligations concerning the Software. > # > # The Initial Developer of the Original Code is: Sun Microsystems, > Inc. > # > # Copyright: 2001 by Sun Microsystems, Inc. > # > # All Rights Reserved. > # > ###################################################################### > #### > #___INFO__MARK_END__ > # > # Template Grid Engine path aliasing configuration file > # > # The following entry aliases physical address as generated by > automounter > # (with a leading /tmp_mnt) to the logical path (w/o leading / > tmp_mnt). > # > # subm_dir subm_host exec_host path_replacement > /tmp_mnt/ * * / > /Volumes/Clinical/ * * / > /export/data/ * * / > > > 10/05/2007 15:25:55 [813:25899]: error: can't > open output file "/Volumes/Clinical/data/lokke_test/ > tky_13dir.bedpostX/logs": No such file or directory > >> Assuming the links are as above, then I think sge_aliases should >> contain "/Volumes/data_directory * * /data_directory", however, we >> re-engineered our setup to avoid these problems, i.e., our / >> Volumes/Data is mounted as /Volumes/Data everywhere, so I'm not >> entirely certain of this. >> >>> when we submit the job via command line with the full path it >>> fails. when we submit the job with the /data_directory path it >>> works. >>> >>> i also want to run FEEDS on this to get more info, but it doesn't >>> seem to submit the job to the cluster and only runs on the head >>> node. is there a way to submit the FEEDS job to the cluster? >> >> The FEEDS script unsets SGE_ROOT at the start. Just comment this >> liine out and replace it with 'exec $FSLTCLSH :$0" "$@"'. It >> should then run the SGE aware subsections using available SGE >> slots. Note that some scripts will then exit immediately leaving >> just their processing tasks on the queues. To work out how long >> everything takes you'll need to manually add up all the runtimes >> (use qacct) and/or observe the wall-clock time data. > > i'll worry about FEEDS on monday, i was trying to run it to give > you guys as much info as possible. > > also, i tried executing the job from the GUI, and after getting all > the data loaded, i cut the /Volumes/Clinical out of the input > directoy box (leaving /data/lokke_test/tky_13dir) and it worked. > so the path is definitely the issue, not the GUI. so why wouldn't > the sge_aliases solve this? > > thanks for the help. > > lokke -- Cheers, Dave Dave Flitney, IT Manager Oxford Centre for Functional MRI of the Brain E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 URL: http://www.fmrib.ox.ac.uk/~flitney --Apple-Mail-4-941695492 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1 Just had a re-read of the docs. = The path aliasing feature appears to be used to resolve problems with = current working directory only, i.e., not general path aliasing (See = page 102 of ng1e6-admin.pdf). If this is the case, then you can't use = this feature to re-map paths in command lines. You'll have to provide = all the required paths or re-factor as we did to make all mount points = the same across our cluster nodes and submission = hosts.
=A0
On 5 Oct 2007, at 20:16, Lokke Highstein = wrote:

On Oct 5, 2007, at 5:45 AM, Dave = Flitney wrote:

=
Hi Lokke,

On 4 Oct = 2007, at 22:11, Lokke Highstein wrote:

one thing is that the paths to the data we are = working on are slightly different on the head nodes and the cluster = nodes, due to the head node having the data mounted off of an xraid (and = showing up in /Volumes, which i then made symbolic links to the / = directory) and the nodes mounting the same directories at /nfs/ with = symbolic links to /

Could you confirm:

On the = head node /data_directory is a link to /Volumes/data_directory
On compute node /data_directory is a link to = /nfs/data_directory

well i was using placeholder = names (data_directory) but the actual data in this case lives in = /Volumes/Clinical/data/lokke_test/tky_13dir on the head node.

also on = the head node /data is a link to /Volumes/Clinical/data

on the = cluster node it's in /nfs/Clinical/data/lokke_test/tky_13dir

and on = that node /data is a link to /nfs/Clinical/data

FYI the = use of the name "Clinical" is just bad naming, it isn't actually = clinical data there, but not it's hard for me to change the name at this = point (long story, not worth telling.)

i have set up sge_aliases which = are supposed to solve this, but still it seems that when we submit a = bedpost job through the GUI (which forces the full path to be used - = even if we type in the /data_directory path it then resolves the actual = full path) it ends up in an error state in the queue.
=

Could you send me the output of: 'cat = ${SGE_ROOT}/default/common/sge_aliases'
Also an = example of the full error message (try 'qstat -j <jobid> -xml | = grep -i QIM_message')


#___INFO__MARK_BEGIN__
#
#=A0 The Contents of this file are = made available subject to the terms of
#=A0 the Sun Industry Standards = Source License Version 1.2
#
#=A0 = Sun Microsystems Inc., March, 2001
#
#=A0 Sun Industry Standards Source = License Version 1.2
#=A0 = =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D
#=A0 The contents of this file are = subject to the Sun Industry Standards
#=A0 Source License Version 1.2 = (the "License"); You may not use this file
#=A0 except in compliance with the = License. You may obtain a copy of the
#=A0 License at htt= p://gridengine.sunsource.net/Gridengine_SISSL_license.html
#
#=A0 Software provided under this = License is provided on an "AS IS" basis,
#=A0 WITHOUT WARRANTY OF ANY KIND, = EITHER EXPRESSED OR IMPLIED, INCLUDING,
#=A0 WITHOUT LIMITATION, = WARRANTIES THAT THE SOFTWARE IS FREE OF DEFECTS,
#=A0 = MERCHANTABLE, FIT FOR A PARTICULAR PURPOSE, OR = NON-INFRINGING.
#=A0 See the License for the = specific provisions governing your rights and
#=A0 = obligations concerning the Software.
#=A0 The Initial Developer of the = Original Code is: Sun Microsystems, Inc.
#=A0 Copyright: 2001 by Sun = Microsystems, Inc.
#
#=A0 = All Rights Reserved.
#
#___INFO__MARK_END__
#
# Template Grid Engine path = aliasing configuration file
#
# The following entry aliases physical address as = generated by automounter
# (with a = leading /tmp_mnt) to the logical path (w/o leading /tmp_mnt).
#
# subm_dir=A0 =A0 =A0 subm_host =A0 =A0 =A0 exec_host =A0 =A0 =A0 = path_replacement
/tmp_mnt/ = =A0 =A0 =A0 * =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 = * =A0 =A0 =A0 =A0 =A0 =A0 =A0= /
/Volumes/Clinical/ * * = /
/export/data/ * * /



Assuming the links are as above, = then I think sge_aliases should contain "/Volumes/data_directory * * = /data_directory", however, we re-engineered our setup to avoid these = problems, i.e., our /Volumes/Data is mounted as /Volumes/Data = everywhere, so I'm not entirely certain of this.

when we submit the job via = command line with the full path it fails.=A0 when we submit the job with = the /data_directory path it works.

i also want to run FEEDS on this = to get more info, but it doesn't seem to submit the job to the cluster = and only runs on the head node.=A0 = is there a way to submit the FEEDS job to the cluster?
=

The FEEDS script unsets SGE_ROOT at the start. Just = comment this liine out and replace it with 'exec $FSLTCLSH :$0" "$@"'. = It should then run the SGE aware subsections using available SGE slots. = Note that some scripts will then exit immediately leaving just their = processing tasks on the queues. To work out how long everything takes = you'll need to manually add up all the runtimes (use qacct) and/or = observe the wall-clock time data.

i'll = worry about FEEDS on monday, i was trying to run it to give you guys as = much info as possible.
also, i tried executing the job = from the GUI, and after getting all the data loaded, i cut the = /Volumes/Clinical out of the input directoy box (leaving = /data/lokke_test/tky_13dir) and it worked.=A0 so the path is definitely the = issue, not the GUI.=A0 so = why wouldn't the sge_aliases solve this?

thanks for = the help.

lokke

Dave Flitney, = IT Manager
Oxford Centre for Functional MRI = of the Brain
E:[log in to unmask] = W:+44-1865-222713 F:+44-1865-222717
URL: http://www.fmrib.ox.ac.uk/~fli= tney

=

= --Apple-Mail-4-941695492-- ========================================================================= Date: Mon, 8 Oct 2007 17:25:07 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mark Jenkinson <[log in to unmask]> Subject: Re: Constrain FLIRT in sch3Dtrans_3dof In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi Brandon, I'm afraid there are no explicit limits in FLIRT anywhere. However, you can limit how far it tries to initialise. In the file sch3Dtrans_3dof, you'll see a whole group of lines that look like: optimise 12 UA:1 0.0 0.0 0.0 0.0 0.0 96.0 0.0 abs 4 The number 96.0 here is the initial translation it tries (here in z, as it is parameter number 6 - but you'll see that this kind of block appears 3 times; once for x, once for y and once for z) If you delete all lines with a translation larger than you think is reasonable then it won't try these. However, it does do a local optimisation from each of these starting positions, so there is no guarantee that it won't go off and still find a ridiculously large translation, but limiting the initialisation like this might help. If not, let us know because it is probably a symptom of something else going wrong. In terms of the schedule file - the best thing is to make a local copy and edit that (by deleting appropriate lines). You then simply give flirt the appropriate path to this local file (as they can be anywhere - flirt doesn't care). Hope this helps. All the best, Mark On 8 Oct 2007, at 11:28, Brandon Whitcher wrote: > Dear FSL experts, > > I am performing a 3D-to-3D registration in FLIRT using the schedule > for > translations only (sch3Dtrans_3dof). > > Question = How can I limit the amount of possible translations in a > given > application of FLIRT? Some of my registrations are producing crazy > translations. > > thank-you, > > Brandon ========================================================================= Date: Mon, 8 Oct 2007 17:46:56 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mark Jenkinson <[log in to unmask]> Subject: Re: applyxfm4d versus applyxfm In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, That's right. There shouldn't be any real difference (although sometimes there are very slight numerical differences, but nothing big) if you only want to apply a single transformation to all the volumes in a 4D series. All the best, Mark On 8 Oct 2007, at 10:45, Lea Krugel wrote: > Dear FSL people, > > I'd like to know whether "applyxfm4D" with only one transformation > matrix > for all the volumes of a functional data set leads to somehow > different > results than flirt in combination with "applyxfm" and "init". > I'm asking because I used "applyxfm" for my functional data and > FEAT has no > problems in recognising the correct number of time points, so I > guess the > "applyxfm" command works fine here, also for functional data. My > understanding of "applyxfm4D" would then be that it is the right > tool when > applying more than one transformation matrix to the data, but with > only one > there is no difference between "applyxfm" and "applyxfm4D". Is that > right or > might there be problems with my output in later analysis steps? > > Thanks a lot, > Lea > > > > ------------------------------------------- > Lea Krugel > > Max Planck Institute for Human Development > Neurocognition of decision making > Lentzeallee 94 > 14195 Berlin > Tel.: +49-30-82406618 > Fax: +49-30-82406616 > email: [log in to unmask] ========================================================================= Date: Mon, 8 Oct 2007 17:54:16 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Christopher Bell <[log in to unmask]> Subject: sigloss Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Hello, I am interested in using sigloss to improve the z-shift registration problems between my structural image and my fmri images. When I try to en= ter the flag --te 4.37 which is the echo time for the phase image, fsl outp= uts "unknown error!" The program seems to work when I do not put a te time in= at all. I also tried using all three slice directions with the same error. H= ow best do I apply the weighted map using flirt? --refweight in an fmri to structural transformation? Thanks for any help you can provide. Chris Bell ========================================================================= Date: Mon, 8 Oct 2007 18:01:08 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mark Jenkinson <[log in to unmask]> Subject: Re: sigloss In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi Chris, This is just a syntax issue. The -- options require a "=" sign. So what you need is: --te=4.37 The single "-" options do not need the "=" sign, and require a space. However, the older FSL programs are (unfortunately) different (such as FLIRT) and always work with spaces, although they do not take "--" options (it is -refweight in FLIRT). As for how to use the sigloss result: you are right that it should be used as a cost function weighting, although usually the EPI is the input volume (as you should use the "nicer" volume as the reference) in which case you want to use the -inweight option. Hope this is helpful. All the best, Mark On 8 Oct 2007, at 17:54, Christopher Bell wrote: > Hello, > > I am interested in using sigloss to improve the z-shift registration > problems between my structural image and my fmri images. When I try > to enter > the flag --te 4.37 which is the echo time for the phase image, > fsl outputs > "unknown error!" The program seems to work when I do not put a te > time in at > all. I also tried using all three slice directions with the same > error. How > best do I apply the weighted map using flirt? --refweight in an > fmri to > structural transformation? Thanks for any help you can provide. > > Chris Bell ========================================================================= Date: Mon, 8 Oct 2007 13:06:57 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Najmeh Khalili M." <[log in to unmask]> Subject: Re: fslview: segmentation fault In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi Dave, I have the fslview (Version 2.4pre0), running on: Debian GNU/Linux 4.0 \n \l rosaline:~# uname -a Linux rosaline 2.6.22.2-i686-smp #1 SMP Tue Sep 25 14:47:07 EDT 2007 i686 GNU/Linux Thanks, Naj On Mon, 8 Oct 2007, Dave Flitney wrote: > Naj, > > It opens fine on the platforms I've tried so far - FSLView 3.0 on: > MacOSX (PPC and Intel); Fedora Core 5 (64bit); and Centos4 (32bit). > > Which OS & FSLView versions are you using? > > On 5 Oct 2007, at 16:51, Najmeh Khalili M. wrote: > > > Hi Dave, > > > > Ref number is: > > 943835 > > > > This is supposed to be a mask. It is not empty at fslstats > > indicates non-zero content. > > Didn't think it was. That was a different poster :-) > > > Thanks > > Naj > > -- > Cheers, Dave > > Dave Flitney, IT Manager > Oxford Centre for Functional MRI of the Brain > E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 > URL: http://www.fmrib.ox.ac.uk/~flitney > > > ========================================================================= Date: Mon, 8 Oct 2007 19:12:50 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Lea Krugel <[log in to unmask]> Subject: Re: applyxfm4d versus applyxfm Comments: To: Mark Jenkinson <[log in to unmask]> In-Reply-To: <[log in to unmask]> Mime-version: 1.0 Content-type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit Hi Mark, thanks a lot for your answer! In the meantime I calculated the difference between the flirt_applyxfm output and the applyxfm4D output, using fslmaths. The resulting difference has a mean intensity that corresponds to about 2 % of the mean intensitity of the original time series (I looked that up in fslview). I also noticed 2 other things: first, the flirt_applyxfm output looks a bit smoother than the applyxfm4d output, and second, the applyxfm4d procedure takes much longer than the flirt_applyxfm operation. The 2 % seem a rather high difference to me, so I now have to decide which of the two possibilities I choose... I have the intuition that the applyxfm4D might be the more exact one because it takes longer to calculate:-)... What would you suggest? Cheers, Lea Am 08.10.2007 18:46 Uhr schrieb "Mark Jenkinson" unter <[log in to unmask]>: > Hi, > > That's right. > There shouldn't be any real difference (although sometimes there > are very slight numerical differences, but nothing big) if you only > want to apply a single transformation to all the volumes in a 4D series. > > All the best, > Mark > > > On 8 Oct 2007, at 10:45, Lea Krugel wrote: > >> Dear FSL people, >> >> I'd like to know whether "applyxfm4D" with only one transformation >> matrix >> for all the volumes of a functional data set leads to somehow >> different >> results than flirt in combination with "applyxfm" and "init". >> I'm asking because I used "applyxfm" for my functional data and >> FEAT has no >> problems in recognising the correct number of time points, so I >> guess the >> "applyxfm" command works fine here, also for functional data. My >> understanding of "applyxfm4D" would then be that it is the right >> tool when >> applying more than one transformation matrix to the data, but with >> only one >> there is no difference between "applyxfm" and "applyxfm4D". Is that >> right or >> might there be problems with my output in later analysis steps? >> >> Thanks a lot, >> Lea >> >> >> >> ------------------------------------------- >> Lea Krugel >> >> Max Planck Institute for Human Development >> Neurocognition of decision making >> Lentzeallee 94 >> 14195 Berlin >> Tel.: +49-30-82406618 >> Fax: +49-30-82406616 >> email: [log in to unmask] ------------------------------------------- Lea Krugel Max Planck Institute for Human Development Neurocognition of decision making Lentzeallee 94 14195 Berlin Tel.: +49-30-82406618 Fax: +49-30-82406616 email: [log in to unmask] ========================================================================= Date: Mon, 8 Oct 2007 18:21:53 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mark Jenkinson <[log in to unmask]> Subject: Re: applyxfm4d versus applyxfm In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Sorry - I should have said that they are the same if you apply sinc interpolation in flirt (with a sincwidth kernel size of 7 - which is the default size). If you just use the defaults in flirt then they are not the same, as flirt will use trilinear which will be smoother (as you've seen). If you smooth your data after this then it doesn't matter than much, although you will be effectively getting a bit less total smoothing from the applyxfm4D output than would typically occur in the FEAT pipeline (which doesn't use sinc). All the best, Mark On 8 Oct 2007, at 18:12, Lea Krugel wrote: > Hi Mark, > > thanks a lot for your answer! > > In the meantime I calculated the difference between the flirt_applyxfm > output and the applyxfm4D output, using fslmaths. The resulting > difference > has a mean intensity that corresponds to about 2 % of the mean > intensitity > of the original time series (I looked that up in fslview). > I also noticed 2 other things: first, the flirt_applyxfm output > looks a bit > smoother than the applyxfm4d output, and second, the applyxfm4d > procedure > takes much longer than the flirt_applyxfm operation. > > The 2 % seem a rather high difference to me, so I now have to > decide which > of the two possibilities I choose... I have the intuition that the > applyxfm4D might be the more exact one because it takes longer to > calculate:-)... What would you suggest? > > Cheers, > Lea > > > Am 08.10.2007 18:46 Uhr schrieb "Mark Jenkinson" unter > <[log in to unmask]>: > >> Hi, >> >> That's right. >> There shouldn't be any real difference (although sometimes there >> are very slight numerical differences, but nothing big) if you only >> want to apply a single transformation to all the volumes in a 4D >> series. >> >> All the best, >> Mark >> >> >> On 8 Oct 2007, at 10:45, Lea Krugel wrote: >> >>> Dear FSL people, >>> >>> I'd like to know whether "applyxfm4D" with only one transformation >>> matrix >>> for all the volumes of a functional data set leads to somehow >>> different >>> results than flirt in combination with "applyxfm" and "init". >>> I'm asking because I used "applyxfm" for my functional data and >>> FEAT has no >>> problems in recognising the correct number of time points, so I >>> guess the >>> "applyxfm" command works fine here, also for functional data. My >>> understanding of "applyxfm4D" would then be that it is the right >>> tool when >>> applying more than one transformation matrix to the data, but with >>> only one >>> there is no difference between "applyxfm" and "applyxfm4D". Is that >>> right or >>> might there be problems with my output in later analysis steps? >>> >>> Thanks a lot, >>> Lea >>> >>> >>> >>> ------------------------------------------- >>> Lea Krugel >>> >>> Max Planck Institute for Human Development >>> Neurocognition of decision making >>> Lentzeallee 94 >>> 14195 Berlin >>> Tel.: +49-30-82406618 >>> Fax: +49-30-82406616 >>> email: [log in to unmask] > > ------------------------------------------- > Lea Krugel > > Max Planck Institute for Human Development > Neurocognition of decision making > Lentzeallee 94 > 14195 Berlin > Tel.: +49-30-82406618 > Fax: +49-30-82406616 > email: [log in to unmask] ========================================================================= Date: Mon, 8 Oct 2007 15:40:32 -0300 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Thomas Doring <[log in to unmask]> Organization: CDPI Subject: Running TBSS on four CPUs Content-Type: text/plain Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Hi, i am trying to run the registration step, tbss_2_reg -T, with all my 4 CPUs but when i open the second terminal i just get this result: [thomas@localhost TBSS]# tbss_2_reg -T wc: .commands: No such file or directory 16862 Somebody has an idea? Thanks, Thomas ========================================================================= Date: Mon, 8 Oct 2007 17:59:38 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Buyean Lee <[log in to unmask]> Subject: FLIRT: Preserver what? In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="--------MB_8C9D80DA824A7B0_698_6524_mblk-d44.sysops.aol.com" ----------MB_8C9D80DA824A7B0_698_6524_mblk-d44.sysops.aol.com Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii" Dear FSL users, When one normalizes a image in SPM2(5), one can choose which parameter (e.g., concentration or total (=amount)) is going to be preserved. I read the FSL website and checked FLIRT options, but I can't find a similar option. Does FLIRT preserve the concentration? Thank you, Buyean ________________________________________________________________________ Check Out the new free AIM(R) Mail -- Unlimited storage and industry-leading spam and email virus protection. ----------MB_8C9D80DA824A7B0_698_6524_mblk-d44.sysops.aol.com Content-Transfer-Encoding: 7bit Content-Type: text/html; charset="us-ascii" Dear FSL users,

When one normalizes a image in SPM2(5), one can choose which parameter (e.g., concentration or total (=amount)) is going to be preserved.

I read the FSL website and checked FLIRT options, but I can't find a similar option.

Does FLIRT preserve the concentration?

Thank you,

Buyean

Check Out the new free AIM(R) Mail -- Unlimited storage and industry-leading spam and email virus protection.
----------MB_8C9D80DA824A7B0_698_6524_mblk-d44.sysops.aol.com-- ========================================================================= Date: Tue, 9 Oct 2007 07:04:59 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: FLIRT: Preserver what? In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, FLIRT applies an affine-only transformation. If you wish to divide the output image by the amount of volumetric scaling (to preserve "concentration") you can extract the volumetric scaling factor using avscale and then apply this easily with the -mul option inside fslmaths. Cheers, Steve. On 8 Oct 2007, at 22:59, Buyean Lee wrote: > Dear FSL users, > > When one normalizes a image in SPM2(5), one can choose which > parameter (e.g., concentration or total (=amount)) is going to be > preserved. > > I read the FSL website and checked FLIRT options, but I can't find > a similar option. > > Does FLIRT preserve the concentration? > > Thank you, > > Buyean > Check Out the new free AIM(R) Mail -- Unlimited storage and > industry-leading spam and email virus protection. ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Tue, 9 Oct 2007 13:40:40 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Lea Krugel <[log in to unmask]> Subject: Re: applyxfm4d versus applyxfm Comments: To: Mark Jenkinson <[log in to unmask]> In-Reply-To: <[log in to unmask]> Mime-version: 1.0 Content-type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit Thank you, Mark, for your helpful explanation! I think I'll take the applyxfm4d output then for the anaylsis. Lea Am 08.10.2007 19:21 Uhr schrieb "Mark Jenkinson" unter <[log in to unmask]>: > Sorry - I should have said that they are the same if > you apply sinc interpolation in flirt (with a sincwidth > kernel size of 7 - which is the default size). If you just > use the defaults in flirt then they are not the same, as > flirt will use trilinear which will be smoother (as you've > seen). If you smooth your data after this then it > doesn't matter than much, although you will be > effectively getting a bit less total smoothing from the > applyxfm4D output than would typically occur in the > FEAT pipeline (which doesn't use sinc). > > All the best, > Mark > > > > > On 8 Oct 2007, at 18:12, Lea Krugel wrote: > >> Hi Mark, >> >> thanks a lot for your answer! >> >> In the meantime I calculated the difference between the flirt_applyxfm >> output and the applyxfm4D output, using fslmaths. The resulting >> difference >> has a mean intensity that corresponds to about 2 % of the mean >> intensitity >> of the original time series (I looked that up in fslview). >> I also noticed 2 other things: first, the flirt_applyxfm output >> looks a bit >> smoother than the applyxfm4d output, and second, the applyxfm4d >> procedure >> takes much longer than the flirt_applyxfm operation. >> >> The 2 % seem a rather high difference to me, so I now have to >> decide which >> of the two possibilities I choose... I have the intuition that the >> applyxfm4D might be the more exact one because it takes longer to >> calculate:-)... What would you suggest? >> >> Cheers, >> Lea >> >> >> Am 08.10.2007 18:46 Uhr schrieb "Mark Jenkinson" unter >> <[log in to unmask]>: >> >>> Hi, >>> >>> That's right. >>> There shouldn't be any real difference (although sometimes there >>> are very slight numerical differences, but nothing big) if you only >>> want to apply a single transformation to all the volumes in a 4D >>> series. >>> >>> All the best, >>> Mark >>> >>> >>> On 8 Oct 2007, at 10:45, Lea Krugel wrote: >>> >>>> Dear FSL people, >>>> >>>> I'd like to know whether "applyxfm4D" with only one transformation >>>> matrix >>>> for all the volumes of a functional data set leads to somehow >>>> different >>>> results than flirt in combination with "applyxfm" and "init". >>>> I'm asking because I used "applyxfm" for my functional data and >>>> FEAT has no >>>> problems in recognising the correct number of time points, so I >>>> guess the >>>> "applyxfm" command works fine here, also for functional data. My >>>> understanding of "applyxfm4D" would then be that it is the right >>>> tool when >>>> applying more than one transformation matrix to the data, but with >>>> only one >>>> there is no difference between "applyxfm" and "applyxfm4D". Is that >>>> right or >>>> might there be problems with my output in later analysis steps? >>>> >>>> Thanks a lot, >>>> Lea >>>> >>>> >>>> >>>> ------------------------------------------- >>>> Lea Krugel >>>> >>>> Max Planck Institute for Human Development >>>> Neurocognition of decision making >>>> Lentzeallee 94 >>>> 14195 Berlin >>>> Tel.: +49-30-82406618 >>>> Fax: +49-30-82406616 >>>> email: [log in to unmask] >> >> ------------------------------------------- >> Lea Krugel >> >> Max Planck Institute for Human Development >> Neurocognition of decision making >> Lentzeallee 94 >> 14195 Berlin >> Tel.: +49-30-82406618 >> Fax: +49-30-82406616 >> email: [log in to unmask] ------------------------------------------- Lea Krugel Max Planck Institute for Human Development Neurocognition of decision making Lentzeallee 94 14195 Berlin Tel.: +49-30-82406618 Fax: +49-30-82406616 email: [log in to unmask] ========================================================================= Date: Tue, 9 Oct 2007 14:16:58 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Benny Liberg <[log in to unmask]> Subject: Re: Time series plot in mask at first and higher level In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain;charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hi, Thanks for advice. It was a very convenient solution. However, my inclusive mask does not seem to apply to 7 of my 12 subjects. Instead of showing the proper mask (located in the ventral lateral premotor cortex), Featquery displays the bottom EPI slice and lacks coordinates. I checked log files and registration results in .feat-dir on first-level and they seemed fine. Any idea on what may have gone wrong? I can upload the files if needed. Kind regards, Benny Liberg > Hi, > > On 3 Oct 2007, at 17:48, Benny Liberg wrote: > >> Hi, >> >> I have a rather simple question. However, I can't seem to figure >> out the answer myself, in >> fslmeants or tsplot. I have an experiment with three conditions (A, >> B, C). I have run a higher-level >> single group analysis (n=12), and have created an inclusive mask >> that contain commonly activated >> vxls in A, B and C. Now I want to investigate the kinetics of the >> signal (on group level and subject >> level) in this mask. >> >> 1. How do I extract the timeseries data within my mask from each >> condition on group level? > > Use the gorup-level standard space mask as the mask input to running > Featquery on _all_ of the first-level analyses (you can setup > multiple analyses in one go in the GUI if they all take the same > input mask). > >> 2. How can I transform my mask in MNI152-space back to native/first- >> level (I noticed that >> Featquery does this if you apply an atlas-mask when running it on >> first-level data)? > > You don't need to, Featquery will do this for you automatically. > > Steve. > > >> >> Any help is much appreciated. Thanks. >> >> Kind regards, >> Benny > > > ------------------------------------------------------------------------ > --- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > ------------------------------------------------------------------------ > --- > ========================================================================= Date: Tue, 9 Oct 2007 15:38:38 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matthew Webster <[log in to unmask]> Subject: Re: TBSS + randomise and f-stats (fwd) In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, It looks like there is a bug in randomise for "all 1's" fstats - in your case it's the first fstat: ( 1 1 1 ). A fix will be in the next patch, but for now if you change the fts file, randomise should generate the other fstats... Matthew > Sure, > > Is this what you want? > > randomise -i all_FA_skeletonised -o test_out -m > mean_FA_skeleton_mask -d bbb.mat -t bbb.con -f > bbb.fts -n 5000 -c 2.5 -F 4 -V > > bbb.con > ===================== > /ContrastName1 group mean > /ContrastName2 variable > /ContrastName3 age > /NumWaves 3 > /NumContrasts 3 > /PPheights 2.304022e-01 6.939000e+01 > 1.557584e+01 > /RequiredEffect 3.610 4.184 3.618 > > /Matrix > 1.000000e+00 0.000000e+00 0.000000e+00 > 0.000000e+00 1.000000e+00 0.000000e+00 > 0.000000e+00 0.000000e+00 1.000000e+00 > > Thanks a lots > Naj > > > > On Sat, 6 Oct 2007, Matthew Webster wrote: > >> Hi, >> Thanks for the fts file - can you upload the full call to >> randomise as >> well - it looks like something is going wrong during the >> input/initialisation stage? >> >> Matthew >>> Opps, how embarrassing :) >>> >>> But I must be still setting the ftest matrix wrong: >>> >>> Loading Data: >>> Data loaded >>> terminate called after throwing an instance of >>> 'NEWMAT::ProgramException' >>> >>> less bbb.fts: >>> >>> /NumWaves 3 >>> /NumContrasts 7 >>> >>> /Matrix >>> 1 1 1 (e.g. intended for y~Intercept+variabe+Age??) >>> 0 1 1 >>> 1 1 0 >>> 1 0 1 >>> 1 0 0 >>> 0 1 0 >>> 0 0 1 >>> >>> Many thanks, >>> Naj >>> >>> >>> >>> On Sat, 6 Oct 2007, Steve Smith wrote: >>> >>>> It sounds like you're trying to feed in the .fsf (general design >>>> setup file) into the f-tests rather than the .fts (f-test contrasts >>>> specification) file? >>>> Cheers. >>>> >>>> >>>> On 5 Oct 2007, at 19:28, Najmeh Khalili M. wrote: >>>> >>>>> >>>>> Hi, >>>>> >>>>> I am interesed in performing f-tests on FA maps, using >>>>> TBSS. I am trying to use randomise with -f option, but it fails: >>>>> >>>>> "bbb.fsfis not a valid vest file >>>>> terminate called after throwing an instance of >>>>> 'RBD_COMMON::BaseException' " >>>>> >>>>> I am attaching a snapshot of the fsf, in case I have completely >>>> misunderstood the instrctions on setting up f-tests! >>>>> >>>>> (EV1 intercept, EV2 a demeaned variable, EV3 age) >>>>> >>>>> Many thanks, >>>>> Naj >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>> >>>> >>>> ------------------------------------------------------------------- >>>> ----- >>>> --- >>>> Stephen M. Smith, Professor of Biomedical Engineering >>>> Associate Director, Oxford University FMRIB Centre >>>> >>>> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >>>> +44 (0) 1865 222726 (fax 222717) >>>> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >>>> ------------------------------------------------------------------- >>>> ----- >>>> --- >> ========================================================================= Date: Tue, 9 Oct 2007 10:48:09 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Najmeh Khalili M." <[log in to unmask]> Subject: Re: TBSS + randomise and f-stats (fwd) In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Thank you! Naj On Tue, 9 Oct 2007, Matthew Webster wrote: > Hi, > It looks like there is a bug in randomise for "all 1's" fstats > - in your case it's the first fstat: ( 1 1 1 ). A fix will be in the > next patch, but for now if you change the fts file, randomise should > generate the other fstats... > > Matthew > > Sure, > > > > Is this what you want? > > > > randomise -i all_FA_skeletonised -o test_out -m > > mean_FA_skeleton_mask -d bbb.mat -t bbb.con -f > > bbb.fts -n 5000 -c 2.5 -F 4 -V > > > > bbb.con > > ===================== > > /ContrastName1 group mean > > /ContrastName2 variable > > /ContrastName3 age > > /NumWaves 3 > > /NumContrasts 3 > > /PPheights 2.304022e-01 6.939000e+01 > > 1.557584e+01 > > /RequiredEffect 3.610 4.184 3.618 > > > > /Matrix > > 1.000000e+00 0.000000e+00 0.000000e+00 > > 0.000000e+00 1.000000e+00 0.000000e+00 > > 0.000000e+00 0.000000e+00 1.000000e+00 > > > > Thanks a lots > > Naj > > > > > > > > On Sat, 6 Oct 2007, Matthew Webster wrote: > > > >> Hi, > >> Thanks for the fts file - can you upload the full call to > >> randomise as > >> well - it looks like something is going wrong during the > >> input/initialisation stage? > >> > >> Matthew > >>> Opps, how embarrassing :) > >>> > >>> But I must be still setting the ftest matrix wrong: > >>> > >>> Loading Data: > >>> Data loaded > >>> terminate called after throwing an instance of > >>> 'NEWMAT::ProgramException' > >>> > >>> less bbb.fts: > >>> > >>> /NumWaves 3 > >>> /NumContrasts 7 > >>> > >>> /Matrix > >>> 1 1 1 (e.g. intended for y~Intercept+variabe+Age??) > >>> 0 1 1 > >>> 1 1 0 > >>> 1 0 1 > >>> 1 0 0 > >>> 0 1 0 > >>> 0 0 1 > >>> > >>> Many thanks, > >>> Naj > >>> > >>> > >>> > >>> On Sat, 6 Oct 2007, Steve Smith wrote: > >>> > >>>> It sounds like you're trying to feed in the .fsf (general design > >>>> setup file) into the f-tests rather than the .fts (f-test contrasts > >>>> specification) file? > >>>> Cheers. > >>>> > >>>> > >>>> On 5 Oct 2007, at 19:28, Najmeh Khalili M. wrote: > >>>> > >>>>> > >>>>> Hi, > >>>>> > >>>>> I am interesed in performing f-tests on FA maps, using > >>>>> TBSS. I am trying to use randomise with -f option, but it fails: > >>>>> > >>>>> "bbb.fsfis not a valid vest file > >>>>> terminate called after throwing an instance of > >>>>> 'RBD_COMMON::BaseException' " > >>>>> > >>>>> I am attaching a snapshot of the fsf, in case I have completely > >>>> misunderstood the instrctions on setting up f-tests! > >>>>> > >>>>> (EV1 intercept, EV2 a demeaned variable, EV3 age) > >>>>> > >>>>> Many thanks, > >>>>> Naj > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>> > >>>> > >>>> ------------------------------------------------------------------- > >>>> ----- > >>>> --- > >>>> Stephen M. Smith, Professor of Biomedical Engineering > >>>> Associate Director, Oxford University FMRIB Centre > >>>> > >>>> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > >>>> +44 (0) 1865 222726 (fax 222717) > >>>> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > >>>> ------------------------------------------------------------------- > >>>> ----- > >>>> --- > >> > ========================================================================= Date: Tue, 9 Oct 2007 23:45:18 +0800 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Yujin Zhang <[log in to unmask]> Subject: Is there the max number of subjects for tbss_4_prestats to project all FA data onto skeleton ? MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_9014_2394893.1191944718762" ------=_Part_9014_2394893.1191944718762 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Dear Sir : I`m working on the DTI data ( 52 subjects ) with FSL-TBSS .It is working well untill running the order "tbss_4_prestats 2000" .It created skeleton mask using threshold 2000 and the skeleton distance map and they looked well ,but it cannot create the all_FA_skeletonised file .It said : projecting all FA data onto skeleton terminate called after throwing an instance of 'std::bad_alloc' what(): St9bad_alloc /usr/local/fsl/bin/tbss_4_prestats: line 99: 11050 Aborted ${FSLDIR}/bin/tbss_skeleton -i mean_FA -p $thresh mean_FA_skeleton_mask_dst ${FSLDIR}/data/standard/LowerCingulum_1mm all_FA all_FA_skeletonised I repeat it 3 times at different computers ,there is not improved . when I reduce the number of subjects to 47 ,it works very well and the all_FA_skeletonised created .Then I try 5, 10, 26 subjects separately ,it works well ,too . Why ? Is there the max number of subjects for tbss_4_prestats to project all FA data onto skeleton ?Can I project two groups subjects (each group has 26 subjects ) FA data onto skeleton separately , and merge them to be all_FA_skeletonised ?There ,skeleton is the skeleton of the mean of all FA data . 5 subjects failed to be normalized when I normalize all FA data using FMRIB58_FA standard space image ,but when I repeated this step with the 5 subjects alone, it succeeded, Do this 5 subjects have some effect to the projection ? Thank you for your reply ! ------=_Part_9014_2394893.1191944718762 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Dear Sir :
I`m working on the DTI data ( 52 subjects ) with FSL-TBSS .It is working well untill running the order "tbss_4_prestats 2000" .It created skeleton mask using threshold 2000 and the skeleton distance map and they looked  well ,but it cannot create the all_FA_skeletonised file .It said :

projecting all FA data onto skeleton
terminate called after throwing an instance of 'std::bad_alloc'
  what():  St9bad_alloc
/usr/local/fsl/bin/tbss_4_prestats: line 99: 11050 Aborted                 ${FSLDIR}/bin/tbss_skeleton -i mean_FA -p $thresh mean_FA_skeleton_mask_dst ${FSLDIR}/data/standard/LowerCingulum_1mm all_FA all_FA_skeletonised

I repeat it 3 times at different computers ,there is not improved .
when I reduce the number of subjects to 47 ,it works very well and the all_FA_skeletonised created .Then I try 5, 10, 26 subjects separately ,it works well ,too .

Why ? Is there the max number of subjects for tbss_4_prestats to project all FA data onto skeleton ?Can I project two groups subjects (each group has 26 subjects ) FA data onto skeleton separately , and merge them to be all_FA_skeletonised ?There ,skeleton is the skeleton of the mean of all FA data .
5 subjects failed to be normalized when I normalize all FA data using FMRIB58_FA standard space image ,but when I repeated this step with the 5 subjects  alone, it  succeeded, Do this 5 subjects have some effect to the projection ?

Thank you for your reply !

------=_Part_9014_2394893.1191944718762-- ========================================================================= Date: Tue, 9 Oct 2007 16:55:04 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: Is there the max number of subjects for tbss_4_prestats to project all FA data onto skeleton ? In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - please see the FAQ: http://www.fmrib.ox.ac.uk/fslfaq/#general_ram Cheers. On 9 Oct 2007, at 16:45, Yujin Zhang wrote: > Dear Sir : > I`m working on the DTI data ( 52 subjects ) with FSL-TBSS .It is > working well untill running the order "tbss_4_prestats 2000" .It > created skeleton mask using threshold 2000 and the skeleton > distance map and they looked well ,but it cannot create the > all_FA_skeletonised file .It said : > > projecting all FA data onto skeleton > terminate called after throwing an instance of 'std::bad_alloc' > what(): St9bad_alloc > /usr/local/fsl/bin/tbss_4_prestats: line 99: 11050 > Aborted ${FSLDIR}/bin/tbss_skeleton -i mean_FA -p > $thresh mean_FA_skeleton_mask_dst ${FSLDIR}/data/standard/ > LowerCingulum_1mm all_FA all_FA_skeletonised > > I repeat it 3 times at different computers ,there is not improved . > when I reduce the number of subjects to 47 ,it works very well and > the all_FA_skeletonised created .Then I try 5, 10, 26 subjects > separately ,it works well ,too . > > Why ? Is there the max number of subjects for tbss_4_prestats to > project all FA data onto skeleton ?Can I project two groups > subjects (each group has 26 subjects ) FA data onto skeleton > separately , and merge them to be all_FA_skeletonised ? > There ,skeleton is the skeleton of the mean of all FA data . > 5 subjects failed to be normalized when I normalize all FA data > using FMRIB58_FA standard space image ,but when I repeated this > step with the 5 subjects alone, it succeeded, Do this 5 subjects > have some effect to the projection ? > > Thank you for your reply ! > ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Tue, 9 Oct 2007 17:40:37 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Bhargav Kumar Errangi <[log in to unmask]> Subject: TBSS-voxelwise correlations between FA and Test scores Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Hello, We are using TBSS to look for voxel-wise correlations between FA and test= scores. Could you please answer the following questions for us? =20 1) Using the GLM tool, we specificy 1 covariate and provide values for ea= ch of 10 subjects in the EV tab. Then in the contrast tab, we specify 2 contrasts (1 and -1) 2) The output is a t statistic image. We were exepcting an r statistic ma= p. Does FSL calculate the t from the r? 3) Will the "1" contrast yield positive correlations and the "-1" contras= t yield negative correlations? 4) The t statistic values are very, very large. We only get reasonbable m= aps when we threshold at t>20. Do you know why this might be? 5) Is there a way to threshold the map based on p-values rather than t statistics? =20 Thank you very much for your assistance. ========================================================================= Date: Tue, 9 Oct 2007 18:02:56 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Teena Moody <[log in to unmask]> Subject: WARNING: posn not header size error Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain I'm getting the following error(and more of the same) in my prestats anal= ysis w fsl 4.0. I=20 converted all data to NIFTI.GZ format. Seems like something is garbled t= hough. Can someone=20 help me decipher this?=20 Thanks Network/Servers/.../bin/fsl/bin/fslroi /.../net_study/s1_1st/raw/net1 exa= mple_func 50 1 ** WARNING: posn not header size (0, 348) ** WARNING: posn not header size (0, 348) log info:=20 ** WARNING: posn not header size (0, 348) 3.1250000000=20 ** WARNING: posn not header size (0, 348) while executing "exec sh -c "$FSLDIR/bin/fslval $thefile pixdim1" " (procedure "feat5:updateimageinfo" line 19) invoked from within "feat5:updateimageinfo $w $i 1" (procedure "feat5:multiple_check" line 43) invoked from within "feat5:multiple_check .r 0 1 1 d" invoked from within ".dialog1.cancel invoke" ("uplevel" body line 1) invoked from within "uplevel #0 [list $w invoke]" (procedure "tk::ButtonUp" line 22) invoked from within "tk::ButtonUp .dialog1.cancel" (command bound to event) ========================================================================= Date: Tue, 9 Oct 2007 18:08:05 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: WARNING: posn not header size error In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - yes, looks like your input data is not valid NIFTI. What do you get with: fslhd net1 fslstats net1 -r -R Cheers. On 9 Oct 2007, at 18:02, Teena Moody wrote: > I'm getting the following error(and more of the same) in my > prestats analysis w fsl 4.0. I > converted all data to NIFTI.GZ format. Seems like something is > garbled though. Can someone > help me decipher this? > Thanks > > Network/Servers/.../bin/fsl/bin/fslroi /.../net_study/s1_1st/raw/ > net1 example_func 50 1 > ** WARNING: posn not header size (0, 348) > ** WARNING: posn not header size (0, 348) > > log info: > ** WARNING: posn not header size (0, 348) > 3.1250000000 > ** WARNING: posn not header size (0, 348) > while executing > "exec sh -c "$FSLDIR/bin/fslval $thefile pixdim1" " > (procedure "feat5:updateimageinfo" line 19) > invoked from within > "feat5:updateimageinfo $w $i 1" > (procedure "feat5:multiple_check" line 43) > invoked from within > "feat5:multiple_check .r 0 1 1 d" > invoked from within > ".dialog1.cancel invoke" > ("uplevel" body line 1) > invoked from within > "uplevel #0 [list $w invoke]" > (procedure "tk::ButtonUp" line 22) > invoked from within > "tk::ButtonUp .dialog1.cancel" > (command bound to event) ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Tue, 9 Oct 2007 18:35:32 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Christopher Petty <[log in to unmask]> Subject: first_utils Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="windows-1252" i have a group of 49 subjects run through first, i have run concat_bvars=20= for the regions that i am interested in. i used Glm to set up the design= =20 matrix as described (groups are broken into 2 EV with a 1 or 0, the group= =20 column is all 1s). I have set up group contrasts as 1 -1. however when i try to run the vertex analysis: first_utils --usebvars --vertexAnalysis -i all_R_Amyg.bvars -d design.mat= - o group_R_amyg --useReconMNI i get this error: terminate called after throwing an instance of 'NEWMAT::SingularException= ' Any ideas? All of my subjects have the corresponding files, and everyone is accounte= d=20 for in the design.mat (49 subjects in bvars and .mat) thanks, -chris ========================================================================= Date: Tue, 9 Oct 2007 19:50:58 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Christopher Petty <[log in to unmask]> Subject: Re: first_utils Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="windows-1252" ok, the preliminary problem .. i needed to use 1 EV with 1 or -1 for the=20= group difference. apparently the group difference needed to be in 1 ev and not multiple evs= ,=20 but how would you add an additional regressor if you can only use 1 EV? ========================================================================= Date: Tue, 9 Oct 2007 20:34:19 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Teena Moody <[log in to unmask]> Subject: Re: WARNING: posn not header size error Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" regarding nifti problem posn error Here's what I get:=20 fslstats net1 -r -R ** WARNING: posn not header size (0, 348) 0.464809 914.370178 -16512.000000 31457.000000=20 fslhd net1.nii.gz=20 ** WARNING: posn not header size (0, 348) filename net1.nii.gz sizeof_hdr 348 data_type INT16 dim0 4 dim1 64 dim2 64 dim3 28 dim4 100 dim5 1 dim6 1 dim7 1 vox_units mm time_units s datatype 4 nbyper 2 bitpix 16 pixdim0 0.0000000000 pixdim1 3.1250000000 pixdim2 3.1250000000 pixdim3 3.9900000095 pixdim4 1.0000000000 pixdim5 0.0000000000 pixdim6 0.0000000000 pixdim7 0.0000000000 vox_offset 352 cal_max 1923.0000 cal_min 0.0000 scl_slope 1.000000 scl_inter 0.000000 phase_dim 0 freq_dim 0 slice_dim 0 slice_name Unknown slice_code 0 slice_start 0 slice_end 0 slice_duration 0.000000 time_offset 0.000000 intent Unknown intent_code 0 intent_name=20=20=20=20 intent_p1 0.000000 intent_p2 0.000000 intent_p3 0.000000 qform_name Aligned Anat qform_code 2 qto_xyz:1 -3.125000 0.000000 -0.000000 96.875000 qto_xyz:2 0.000000 3.125000 -0.000000 -96.875000 qto_xyz:3 0.000000 0.000000 3.990000 -51.869999 qto_xyz:4 0.000000 0.000000 0.000000 1.000000 qform_xorient Right-to-Left qform_yorient Posterior-to-Anterior qform_zorient Inferior-to-Superior sform_name Aligned Anat sform_code 2 sto_xyz:1 -3.125000 0.000000 0.000000 96.875000 sto_xyz:2 0.000000 3.125000 0.000000 -96.875000 sto_xyz:3 0.000000 0.000000 3.990000 -51.869999 sto_xyz:4 0.000000 0.000000 0.000000 1.000000 sform_xorient Right-to-Left sform_yorient Posterior-to-Anterior sform_zorient Inferior-to-Superior file_type NIFTI-1+ file_code 1 descrip FSL4.0 aux_file=20=20 If I do fslhd on the analyze version of the file, I get:=20 fslhd net_func_axial_6 filename net_func_axial_6 sizeof_hdr 348 data_type dsr db_name net_func_axial_6 extents 16384 session_error 0 regular r hkey_un0=20=20=20=20=20=20=20 dim0 4 dim1 64 dim2 64 dim3 28 dim4 100 dim5 0 dim6 0 dim7 0 vox_units mm cal_units=20=20=20=20=20=20 unused1 0 datatype 4 bitpix 16 pixdim0 4.0000000000 pixdim1 3.1250000000 pixdim2 3.1250000000 pixdim3 3.9900000095 pixdim4 1.0000000000 pixdim5 0.0000000000 pixdim6 0.0000000000 pixdim7 0.0000000000 vox_offset 0.0000 funused1 1.0000 funused2 0.0000 funused3 0.0000 cal_max 1963.0000 cal_min 0.0000 compressed 0 verified 0 glmax 1963 glmin 0 descrip Bookheimer^NET aux_file=20=20=20=20=20=20=20 orient 0 originator=20=20=20=20=20 origin1 0 origin2 0 origin3 0 generated (X)MedCon scannum 1=20 patient_id net_s1_1 exp_date=20=20=20=20=20=20=20 exp_time=20=20=20=20=20=20=20 hist_un0=20=20=20=20=20=20=20 views 0 vols_added 0 start_field 0 field_skip 0 omax 0 omin 0 smin 0 smin 0 file_type ANALYZE-7.5 file_code 0 ========================================================================= Date: Tue, 9 Oct 2007 12:17:08 -0700 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Jeremy Bronson <[log in to unmask]> Subject: Re: FSL on XGrid w/ XServe (now SGE) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: quoted-printable Hi,=0A=0AI tried to get FSL working with XGrid during the summer, but wasn'= t=0A quite successful, and was recommended to use SGE in its place by Steve= and=0A others. I'm giving it another go, but I've run into a few question= s=0A while setting up SGE. Hopefully they're easy enough to answer...=0A=0A= -The SGE manual says that each host must have the same user account=0A name= s and passwords, which is the alternate to XGrid's nobody:nobody=0A permiss= ions, but seems equally impractical. If SGE is really designed to be=0A im= plemented in large grids, I'm not sure how it could scale beyond a=0A handf= ul of managed machines. My question is whether this is how the=0A lab is = configured at Oxford or any other site, or alternately are all=0A jobs subm= itted under a single username?=0A=0A-All scan data resides on an XServe, au= tomounted via NFS. What kind of=0A permissions are necessary on this share?= =0A=0A-Is anyone using the multiple queues available in SGE, or only using= =0A one for large jobs?=0A=0AMy basic goal is to allow users to login, star= t multiple large jobs on=0A the grid, then log out and retrieve the results= later. The FEAT=0A first-level analyses tend to tie up all the machines f= or a time, so I'd love=0A to speed them up by running the jobs in parallel = and in the background.=0A=0A=0AThanks in advance!=0A=0AJeremy=0A=0A=0A=0A= =0A=0A=0AOn Tue, 14 Aug 2007 06:24:00 +0100, Steve Smith <[log in to unmask] .UK>=0A wrote:=0A> Hi Jeremy,=0A> =0A> I agree with Andrew, most people see= m much happier with SGE than with=0A =0A> XGrid, so if it's not too late I= would consider that.=0A> =0A> Anyway - yes, hopefully FSL 4.0 should be fa= irly easy to setup with =0A> either (though much easier with SGE as that's= what we have so should=0A =0A> need much less customising).=0A> =0A> Firs= t, see the brief intro at:=0A> http://www.fmrib.ox.ac.uk/fsl/fsl/downloadin= g.html#sge=0A> So if the user runs any of these programs on a machine which= can =0A> submit to the cluster then that should happen automatically, aft= er =0A> the sysadmin has:=0A> =0A> Setup the central cluster-controlling s= cript=0A> $FSLDIR/bin/fsl_sub=0A> =0A> This is a heavily commented shell sc= ript and hopefully should be =0A> reasonably easy to follow and customise.= ....=0A> =0A> =0A> So hopefully the user can just run FSL programs on their= normal =0A> working machine, and if it's setup to be a cluster submission= host, =0A> then whenever a "big" job is run that will automatically get s= ent to=0A =0A> the queue.=0A> =0A> Cheers, Steve.=0A> =0A> =0A> =0A> =0A> = =0A> On 14 Aug 2007, at 05:59, Jeremy Bronson wrote:=0A> =0A> > Hi All,=0A>= >=0A> > I'm the new administrator of a neuroimaging research lab, and I'm = =0A> > working on getting FSL and other MRI-analysis tools running on =0A= > > XGrid. I'm not yet intimately familiar with how FSL works, so I'm=0A = =0A> > hoping there are others out there who have already figured out how= =0A =0A> > to run it on a cluster, so I don't have to reinvent the wheel. = =0A> > I've heard of several existing clusters that run FSL, albeit a = =0A> > modified version. I'm hoping it can be done with the off-the-shelf= =0A =0A> > version, perhaps it's now possible with the just-released 4.0? = If=0A =0A> > anybody could point me in the direction of some specifics, I= 'd be =0A> > most grateful.=0A> >=0A> > We've got an XServe with RAID that= houses all the data, and FSL is=0A =0A> > installed and configured on all= host (agent) machines. Most users=0A =0A> > use the GUI, and manually po= int the tools to the appropriate data,=0A =0A> > so I assume they'll need = to familiarize themselves with the=0A command- =0A> > line tools and specif= ying data directories on the CLI (or via =0A> > GridStuffer). I'm thinkin= g that each machine will need the data =0A> > volume to be auto-mounted (N= FS?) at startup with appropriate =0A> > directories having read/write acce= ss for the 'nobody' user. Does =0A> > this sound correct?=0A> >=0A> > Addi= tionally, each tool that's part of the FSL package seems to =0A> > launch = a number of other UNIX commands during analysis, to copy, =0A> > move and = otherwise manipulate the result data. Will this confuse =0A> > XGrid, or = will the job and all sub-commands run until the original=0A =0A> > command= completes? (e.g. A complete FEAT analysis)=0A> >=0A> > Hopefully this isn'= t too difficult, and afterwards I'd like to take=0A =0A> > the time to pos= t a HOWTO on macresearch.org or the like, so others=0A =0A> > might take a= dvantage of the info. Thanks in advance to anyone who=0A =0A> > might be = able to help.=0A> >=0A> >=0A> > Jeremy Bronson=0A> > Systems Administrator= =0A> > Frey Research Lab=0A> > University of Oregon=0A> =0A> =0A>=0A ------= ------------------------------------------------------------------ =0A> ---= =0A> Stephen M. Smith, Professor of Biomedical Engineering=0A> Associate Di= rector, Oxford University FMRIB Centre=0A> =0A> FMRIB, JR Hospital, Headin= gton, Oxford OX3 9DU, UK=0A> +44 (0) 1865 222726 (fax 222717)=0A> steve@f= mrib.ox.ac.uk http://www.fmrib.ox.ac.uk/~steve=0A>=0A ------------------= ------------------------------------------------------ =0A> ---=0A> =0A=0A= =0A=0A=0A=0AHi,=0A=0AI tried to get FSL working with XGrid during the summe= r, but wasn't=0A quite successful, and was recommended to use SGE in its pl= ace by Steve and=0A others. I'm giving it another go, but I've run into a = few questions=0A while setting up SGE. Hopefully they're easy enough to ans= wer...=0A=0A-The SGE manual says that each host must have the same user acc= ount=0A names and passwords, which is the alternate to XGrid's nobody:nobod= y=0A permissions, but seems equally impractical. If SGE is really designed= to be=0A implemented in large grids, I'm not sure how it could scale beyon= d a=0A handful of managed machines. My question is whether this is how th= e=0A lab is configured at Oxford or any other site, or alternately are all= =0A jobs submitted under a single username?=0A=0A-All scan data resides on = an XServe, automounted via NFS. What kind of=0A permissions are necessary o= n this share?=0A=0A-Is anyone using the multiple queues available in SGE, o= r only using=0A one for large jobs?=0A=0AMy basic goal is to allow users to= login, start multiple large jobs on=0A the grid, then log out and retrieve= the results later. The FEAT=0A first-level analyses tend to tie up all th= e machines for a time, so I'd love=0A to speed them up by running the jobs = in parallel and in the background.=0A=0A=0AThanks in advance!=0A=0AJeremy= =0A=0A=0A=0A=0A=0A=0AOn Tue, 14 Aug 2007 06:24:00 +0100, Steve Smith [log in to unmask]>=0A wrote:=0A> Hi Jeremy,=0A> =0A> I agree with Andrew, mos= t people seem much happier with SGE than with=0A =0A> XGrid, so if it's no= t too late I would consider that.=0A> =0A> Anyway - yes, hopefully FSL 4.0 = should be fairly easy to setup with =0A> either (though much easier with S= GE as that's what we have so should=0A =0A> need much less customising).= =0A> =0A> First, see the brief intro at:=0A> http://www.fmrib.ox.ac.uk/fsl/= fsl/downloading.html#sge=0A> So if the user runs any of these programs on a= machine which can =0A> submit to the cluster then that should happen auto= matically, after =0A> the sysadmin has:=0A> =0A> Setup the central cluster= -controlling script=0A> $FSLDIR/bin/fsl_sub=0A> =0A> This is a heavily comm= ented shell script and hopefully should be =0A> reasonably easy to follow = and customise.....=0A> =0A> =0A> So hopefully the user can just run FSL pro= grams on their normal =0A> working machine, and if it's setup to be a clus= ter submission host, =0A> then whenever a "big" job is run that will autom= atically get sent to=0A =0A> the queue.=0A> =0A> Cheers, Steve.=0A> =0A> = =0A> =0A> =0A> =0A> On 14 Aug 2007, at 05:59, Jeremy Bronson wrote:=0A> =0A= > > Hi All,=0A> >=0A> > I'm the new administrator of a neuroimaging researc= h lab, and I'm =0A> > working on getting FSL and other MRI-analysis tools = running on =0A> > XGrid. I'm not yet intimately familiar with how FSL wor= ks, so I'm=0A =0A> > hoping there are others out there who have already fi= gured out how=0A =0A> > to run it on a cluster, so I don't have to reinven= t the wheel. =0A> > I've heard of several existing clusters that run FSL,= albeit a =0A> > modified version. I'm hoping it can be done with the off= -the-shelf=0A =0A> > version, perhaps it's now possible with the just-rele= ased 4.0? If=0A =0A> > anybody could point me in the direction of some sp= ecifics, I'd be =0A> > most grateful.=0A> >=0A> > We've got an XServe with= RAID that houses all the data, and FSL is=0A =0A> > installed and configu= red on all host (agent) machines. Most users=0A =0A> > use the GUI, and m= anually point the tools to the appropriate data,=0A =0A> > so I assume the= y'll need to familiarize themselves with the=0A command- =0A> > line tools = and specifying data directories on the CLI (or via =0A> > GridStuffer). I= 'm thinking that each machine will need the data =0A> > volume to be auto-= mounted (NFS?) at startup with appropriate =0A> > directories having read/= write access for the 'nobody' user. Does =0A> > this sound correct?=0A> >= =0A> > Additionally, each tool that's part of the FSL package seems to =0A= > > launch a number of other UNIX commands during analysis, to copy, =0A> = > move and otherwise manipulate the result data. Will this confuse =0A> >= XGrid, or will the job and all sub-commands run until the original=0A =0A= > > command completes? (e.g. A complete FEAT analysis)=0A> >=0A> > Hopefull= y this isn't too difficult, and afterwards I'd like to take=0A =0A> > the = time to post a HOWTO on macresearch.org or the like, so others=0A =0A> > m= ight take advantage of the info. Thanks in advance to anyone who=0A =0A> = > might be able to help.=0A> >=0A> >=0A> > Jeremy Bronson=0A> > Systems Adm= inistrator=0A> > Frey Research Lab=0A> > University of Oregon=0A> =0A> =0A>= =0A -----------------------------------------------------------------------= - =0A> ---=0A> Stephen M. Smith, Professor of Biomedical Engineering=0A> As= sociate Director, Oxford University FMRIB Centre=0A> =0A> FMRIB, JR Hospit= al, Headington, Oxford OX3 9DU, UK=0A> +44 (0) 1865 222726 (fax 222717)= =0A> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve=0A>=0A ------= ------------------------------------------------------------------ =0A> ---= =0A> =0A=0A=0A=0A=0A=0A ========================================================================= Date: Tue, 9 Oct 2007 16:24:28 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Lokke Highstein <[log in to unmask]> Subject: probtrackx GUI issue on OSX cluster Mime-Version: 1.0 (Apple Message framework v752.2) Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed hello list, sorry to keep bugging you all, but we are making a lot of progress with our cluster, thanks to this list, and we are almost fully functional which is very exciting, as we are seeing some very impressive speeds on the larger jobs (especially bedpost!) i implemented the symbolic links which were suggested to me in earlier posts, and now we can run bedpostx jobs and FEAT jobs fine from the gui, however we seem to have hit a snag with probtrackx. here is the information i got from the user in question, please advise what the problem could be, as it works fine from the command line (although for some reason this job gets run on the head node only and never gets submitted through the qsub process to the cluster.) Running probtrackx from the GUI creates the appropriate output directory, but only contains fdt_log.tcl and fdt_script.sh If I open fdt_script.sh and paste only the probtrackx command into the terminal command line, the job runs perfectly. This is the command that I paste: /common/fsl/bin/probtrackx --mode=seedmask -x /Volumes/Clinical/homes/ tyanagihara/sweat/DTI/mask-dump/seed-internalCapsule-mask.hdr -l -c 0.2 -S 2000 --steplength=0.5 -P 5000 --stop=/Volumes/Clinical/homes/ tyanagihara/sweat/DTI/mask-dump/target1-mask.hdr --forcedir --opd -s This is the error I get when running from the GUI: usage: at [-q x] [-f file] [-m] time at -c job [job ...] at [-f file] -t [[CC]YY]MMDDhhmm[.SS] at -r job [job ...] at -l -q queuename at -l [job ...] atq [-q x] [-v] atrm job [job ...] batch [-f file] [-m] usage: at [-q x] [-f file] [-m] time at -c job [job ...] at [-f file] -t [[CC]YY]MMDDhhmm[.SS] at -r job [job ...] at -l -q queuename at -l [job ...] atq [-q x] [-v] atrm job [job ...] batch [-f file] [-m] while executing "exec batch -q long.q ${filebase}_script.sh" ("probtrackx" arm line 144) invoked from within "switch -- $probtrack(tool) { eddy_current { global eddy set errorStr "" if { $eddy(input) == "" } { set errorStr "You need to specify..." (procedure "fdt:apply" line 5) invoked from within "fdt:apply .fdt keep" invoked from within ".fdt.apply invoke" ("uplevel" body line 1) invoked from within "uplevel #0 [list $w invoke]" (procedure "tk::ButtonUp" line 22) invoked from within "tk::ButtonUp .fdt.apply" (command bound to event) ========================================================================= Date: Tue, 9 Oct 2007 19:31:50 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Lokke Highstein <[log in to unmask]> Subject: Re: FSL on XGrid w/ XServe (now SGE) In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit hi jeremy, i'm very new to this but i can already answer one of your questions and then defer to the others here on the second two. we have a 8 node xserve cluster and we use open directory as our master username/password database management system. just about anything that can access LDAP data can authenticate to it so most of the computers (servers included) can share the same username/password database (although the windows xp boxes give me trouble.) open directory is built into OSX server, and on the xserve nodes it's very simple to set up so they authenticate to that system. i'm not sure about permissions, i am still playing with them at the moment, and i also am just starting to look into multiple queues. i'm sure someone else will have more info. -lokke On Oct 9, 2007, at 3:17 PM, Jeremy Bronson wrote: > Hi, > > I tried to get FSL working with XGrid during the summer, but wasn't > quite successful, and was recommended to use SGE in its place by > Steve and > others. I'm giving it another go, but I've run into a few questions > while setting up SGE. Hopefully they're easy enough to answer... > > -The SGE manual says that each host must have the same user account > names and passwords, which is the alternate to XGrid's nobody:nobody > permissions, but seems equally impractical. If SGE is really > designed to be > implemented in large grids, I'm not sure how it could scale beyond a > handful of managed machines. My question is whether this is how the > lab is configured at Oxford or any other site, or alternately are all > jobs submitted under a single username? > > -All scan data resides on an XServe, automounted via NFS. What kind of > permissions are necessary on this share? > > -Is anyone using the multiple queues available in SGE, or only using > one for large jobs? > > My basic goal is to allow users to login, start multiple large jobs on > the grid, then log out and retrieve the results later. The FEAT > first-level analyses tend to tie up all the machines for a time, > so I'd love > to speed them up by running the jobs in parallel and in the > background. > > > Thanks in advance! > > Jeremy > > > > > > > On Tue, 14 Aug 2007 06:24:00 +0100, Steve Smith <[log in to unmask]> > wrote: >> Hi Jeremy, >> >> I agree with Andrew, most people seem much happier with SGE than with > >> XGrid, so if it's not too late I would consider that. >> >> Anyway - yes, hopefully FSL 4.0 should be fairly easy to setup with >> either (though much easier with SGE as that's what we have so should > >> need much less customising). >> >> First, see the brief intro at: >> http://www.fmrib.ox.ac.uk/fsl/fsl/downloading.html#sge >> So if the user runs any of these programs on a machine which can >> submit to the cluster then that should happen automatically, after >> the sysadmin has: >> >> Setup the central cluster-controlling script >> $FSLDIR/bin/fsl_sub >> >> This is a heavily commented shell script and hopefully should be >> reasonably easy to follow and customise..... >> >> >> So hopefully the user can just run FSL programs on their normal >> working machine, and if it's setup to be a cluster submission host, >> then whenever a "big" job is run that will automatically get sent to > >> the queue. >> >> Cheers, Steve. >> >> >> >> >> >> On 14 Aug 2007, at 05:59, Jeremy Bronson wrote: >> >>> Hi All, >>> >>> I'm the new administrator of a neuroimaging research lab, and I'm >>> working on getting FSL and other MRI-analysis tools running on >>> XGrid. I'm not yet intimately familiar with how FSL works, so I'm > >>> hoping there are others out there who have already figured out how > >>> to run it on a cluster, so I don't have to reinvent the wheel. >>> I've heard of several existing clusters that run FSL, albeit a >>> modified version. I'm hoping it can be done with the off-the-shelf > >>> version, perhaps it's now possible with the just-released 4.0? If > >>> anybody could point me in the direction of some specifics, I'd be >>> most grateful. >>> >>> We've got an XServe with RAID that houses all the data, and FSL is > >>> installed and configured on all host (agent) machines. Most users > >>> use the GUI, and manually point the tools to the appropriate data, > >>> so I assume they'll need to familiarize themselves with the > command- >>> line tools and specifying data directories on the CLI (or via >>> GridStuffer). I'm thinking that each machine will need the data >>> volume to be auto-mounted (NFS?) at startup with appropriate >>> directories having read/write access for the 'nobody' user. Does >>> this sound correct? >>> >>> Additionally, each tool that's part of the FSL package seems to >>> launch a number of other UNIX commands during analysis, to copy, >>> move and otherwise manipulate the result data. Will this confuse >>> XGrid, or will the job and all sub-commands run until the original > >>> command completes? (e.g. A complete FEAT analysis) >>> >>> Hopefully this isn't too difficult, and afterwards I'd like to take > >>> the time to post a HOWTO on macresearch.org or the like, so others > >>> might take advantage of the info. Thanks in advance to anyone who > >>> might be able to help. >>> >>> >>> Jeremy Bronson >>> Systems Administrator >>> Frey Research Lab >>> University of Oregon >> >> >> > > ---------------------------------------------------------------------- > -- >> --- >> Stephen M. Smith, Professor of Biomedical Engineering >> Associate Director, Oxford University FMRIB Centre >> >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >> +44 (0) 1865 222726 (fax 222717) >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >> > > ---------------------------------------------------------------------- > -- >> --- >> > > > > > > Hi, > > I tried to get FSL working with XGrid during the summer, but wasn't > quite successful, and was recommended to use SGE in its place by > Steve and > others. I'm giving it another go, but I've run into a few questions > while setting up SGE. Hopefully they're easy enough to answer... > > -The SGE manual says that each host must have the same user account > names and passwords, which is the alternate to XGrid's nobody:nobody > permissions, but seems equally impractical. If SGE is really > designed to be > implemented in large grids, I'm not sure how it could scale beyond a > handful of managed machines. My question is whether this is how the > lab is configured at Oxford or any other site, or alternately are all > jobs submitted under a single username? > > -All scan data resides on an XServe, automounted via NFS. What kind of > permissions are necessary on this share? > > -Is anyone using the multiple queues available in SGE, or only using > one for large jobs? > > My basic goal is to allow users to login, start multiple large jobs on > the grid, then log out and retrieve the results later. The FEAT > first-level analyses tend to tie up all the machines for a time, > so I'd love > to speed them up by running the jobs in parallel and in the > background. > > > Thanks in advance! > > Jeremy > > > > > > > On Tue, 14 Aug 2007 06:24:00 +0100, Steve Smith <[log in to unmask]> > wrote: >> Hi Jeremy, >> >> I agree with Andrew, most people seem much happier with SGE than with > >> XGrid, so if it's not too late I would consider that. >> >> Anyway - yes, hopefully FSL 4.0 should be fairly easy to setup with >> either (though much easier with SGE as that's what we have so should > >> need much less customising). >> >> First, see the brief intro at: >> http://www.fmrib.ox.ac.uk/fsl/fsl/downloading.html#sge >> So if the user runs any of these programs on a machine which can >> submit to the cluster then that should happen automatically, after >> the sysadmin has: >> >> Setup the central cluster-controlling script >> $FSLDIR/bin/fsl_sub >> >> This is a heavily commented shell script and hopefully should be >> reasonably easy to follow and customise..... >> >> >> So hopefully the user can just run FSL programs on their normal >> working machine, and if it's setup to be a cluster submission host, >> then whenever a "big" job is run that will automatically get sent to > >> the queue. >> >> Cheers, Steve. >> >> >> >> >> >> On 14 Aug 2007, at 05:59, Jeremy Bronson wrote: >> >>> Hi All, >>> >>> I'm the new administrator of a neuroimaging research lab, and I'm >>> working on getting FSL and other MRI-analysis tools running on >>> XGrid. I'm not yet intimately familiar with how FSL works, so I'm > >>> hoping there are others out there who have already figured out how > >>> to run it on a cluster, so I don't have to reinvent the wheel. >>> I've heard of several existing clusters that run FSL, albeit a >>> modified version. I'm hoping it can be done with the off-the-shelf > >>> version, perhaps it's now possible with the just-released 4.0? If > >>> anybody could point me in the direction of some specifics, I'd be >>> most grateful. >>> >>> We've got an XServe with RAID that houses all the data, and FSL is > >>> installed and configured on all host (agent) machines. Most users > >>> use the GUI, and manually point the tools to the appropriate data, > >>> so I assume they'll need to familiarize themselves with the > command- >>> line tools and specifying data directories on the CLI (or via >>> GridStuffer). I'm thinking that each machine will need the data >>> volume to be auto-mounted (NFS?) at startup with appropriate >>> directories having read/write access for the 'nobody' user. Does >>> this sound correct? >>> >>> Additionally, each tool that's part of the FSL package seems to >>> launch a number of other UNIX commands during analysis, to copy, >>> move and otherwise manipulate the result data. Will this confuse >>> XGrid, or will the job and all sub-commands run until the original > >>> command completes? (e.g. A complete FEAT analysis) >>> >>> Hopefully this isn't too difficult, and afterwards I'd like to take > >>> the time to post a HOWTO on macresearch.org or the like, so others > >>> might take advantage of the info. Thanks in advance to anyone who > >>> might be able to help. >>> >>> >>> Jeremy Bronson >>> Systems Administrator >>> Frey Research Lab >>> University of Oregon >> >> >> > > ---------------------------------------------------------------------- > -- >> --- >> Stephen M. Smith, Professor of Biomedical Engineering >> Associate Director, Oxford University FMRIB Centre >> >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >> +44 (0) 1865 222726 (fax 222717) >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >> > > ---------------------------------------------------------------------- > -- >> --- >> > > > > > > ========================================================================= Date: Wed, 10 Oct 2007 01:23:40 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Roman Loonis <[log in to unmask]> Subject: Installation worked...then failed Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain I installed FSL on my apple computer at the lab. Everything worked very w= ell until I removed the=20 laptop from the lab and the network at the lab. When I attempted using it= at home, it stopped=20 working... I get the following address: anne-sophie-champods-computer:~ debas$ fsl Application initialization failed: couldn't connect to display "localhost= :11" Error in startup script: couldn't connect to display "localhost:11" while executing "load /usr/local/fsl/extras/lib/tk8.4/../libtk8.4.dylib Tk" ("package ifneeded" script) invoked from within "package require Tk" invoked from within "if { [ string match -nocase *wish* $MYSHELL ] } { package require Tk # tk_focusFollowsMouse #bind Button { focus %W ; tk::ButtonEnter ..." (file "/usr/local/fsl/tcl/fslstart.tcl" line 19) invoked from within "source [ file dirname [ info script ] ]/fslstart.tcl" (file "/usr/local/fsl/tcl/fsl.tcl" line 71) invoked from within "source ${FSLDIR}/tcl/${origname}.tcl" (file "/usr/local/fsl/bin/fsl" line 22) Hopefully someone could help. Thank you,=20 Roman ========================================================================= Date: Wed, 10 Oct 2007 09:07:32 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Colm G. Connolly" <[log in to unmask]> Subject: Re: FSL on XGrid w/ XServe (now SGE) In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=ISO-8859-1; delsp=yes; format=flowed Content-Transfer-Encoding: quoted-printable On 10 Oct 2007, at 00:31, Lokke Highstein wrote: > hi jeremy, > > i'm very new to this but i can already answer one of your questions =20= > and then defer to the others here on the second two. > > we have a 8 node xserve cluster and we use open directory as our =20 > master username/password database management system. just about =20 > anything that can access LDAP data can authenticate to it so most =20 > of the computers (servers included) can share the same username/=20 > password database (although the windows xp boxes give me trouble.) This is definitely the way to do it. We run fsl on a cluster of 356 =20 dual-node AMD Opterons and have all user information stored in an =20 LDAP DB against which all login nodes and client nodes authenticate. =20 In an environment such as this and yours this is best way to do it. =20 As for xp you could set up domains in samba, have it authenticate =20 against the ldap db and use the same username and passwords to login =20 to both the unix and windows machines. It is a lot more complicated =20 that that but it is doable. If you go down this route I recommend =20 that you integrate the samba schema into your ldap db from the =20 outset, it simplifies things greatly. > > open directory is built into OSX server, and on the xserve nodes =20 > it's very simple to set up so they authenticate to that system. > > i'm not sure about permissions, i am still playing with them at the =20= > moment, and i also am just starting to look into multiple queues. I think as long as all clients and servers share the same password db =20= the nfs servers should only need to export your filesystems as r/w so =20= you can both read and write data. Of course your clients will need to =20= mount the exported f/s as r/w. Bye, -- Dr Colm G. Connolly School of Psychology and Institute of Neuroscience The Lloyd Building University of Dublin Trinity College, Dublin 2, =C9ire Tel: +353-1-896-8475 Fax: +353-1-671-3183 ========================================================================= Date: Wed, 10 Oct 2007 09:15:03 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Brian Patenaude <[log in to unmask]> Subject: Re: first_utils In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hi, The restriction is that no linear combination of the columns sum to a column of ones. first_utils demeans the design matrix such that the first column of the design matrix is all ones. This is why splitting the groups into into multiple columns causes an error. first_utils will demean all regressors internally. Cheers, Brian > ok, the preliminary problem .. i needed to use 1 EV with 1 or -1 for the > group difference. > > apparently the group difference needed to be in 1 ev and not multiple > evs, but how would you add an additional regressor if you can only use > 1 EV? ========================================================================= Date: Wed, 10 Oct 2007 10:10:17 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Thomas Nichols <[log in to unmask]> Subject: Re: TBSS-voxelwise correlations between FA and Test scores In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_20113_7950987.1192007417532" ------=_Part_20113_7950987.1192007417532 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Bhargav, Did you include an intercept in the model? Or specify the -D option and use a centered covariate? That might well explain the large t values (without an intercept you may be testing if the FA values are non-zero, not a very interesting question). randomise uses the general linear model and contrasts to test effects of interest with t-tests and so doesn't produce correlation coefficients. You can threshold on the basis of corrected or uncorrected P-values with fslmaths/avwmaths, the -thr option, and the corresponding One-Minus-P-value images; see the randomise help page http://www.fmrib.ox.ac.uk/fsl/randomise/ for the exact filename extensions for each of these types of images. Hope this helps! -Tom On 10/9/07, Bhargav Kumar Errangi <[log in to unmask]> wrote: > > Hello, > > We are using TBSS to look for voxel-wise correlations between FA and test > scores. Could you please answer the following questions for us? > > 1) Using the GLM tool, we specificy 1 covariate and provide values for > each > of 10 subjects in the EV tab. Then in the contrast tab, we specify 2 > contrasts (1 and -1) > 2) The output is a t statistic image. We were exepcting an r statistic > map. > Does FSL calculate the t from the r? > 3) Will the "1" contrast yield positive correlations and the "-1" contrast > yield negative correlations? > 4) The t statistic values are very, very large. We only get reasonbable > maps > when we threshold at t>20. Do you know why this might be? > 5) Is there a way to threshold the map based on p-values rather than t > statistics? > ____________________________________________ Thomas Nichols, PhD Director, Modelling & Genetics GlaxoSmithKline Clinical Imaging Centre Senior Research Fellow Oxford University FMRIB Centre ------=_Part_20113_7950987.1192007417532 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Bhargav,

Did you include an intercept in the model?  Or specify the -D option and use a centered covariate?  That might well explain the large t values (without an intercept you may be testing if the FA values are non-zero, not a very interesting question).

randomise uses the general linear model and contrasts to test effects of interest with t-tests and so doesn't produce correlation coefficients.

You can threshold on the basis of corrected or uncorrected P-values with fslmaths/avwmaths, the -thr option, and the corresponding One-Minus-P-value images;  see the randomise help page http://www.fmrib.ox.ac.uk/fsl/randomise/ for the exact filename extensions for each of these types of images.

Hope this helps!

-Tom

On 10/9/07, Bhargav Kumar Errangi <[log in to unmask]> wrote:
Hello,

We are using TBSS to look for voxel-wise correlations between FA and test
scores. Could you please answer the following questions for us?

1) Using the GLM tool, we specificy 1 covariate and provide values for each
of 10 subjects in the EV tab. Then in the contrast tab, we specify 2
contrasts (1 and -1)
2) The output is a t statistic image. We were exepcting an r statistic map.
Does FSL calculate the t from the r?
3) Will the "1" contrast yield positive correlations and the "-1" contrast
yield negative correlations?
4) The t statistic values are very, very large. We only get reasonbable maps
when we threshold at t>20. Do you know why this might be?
5) Is there a way to threshold the map based on p-values rather than t
statistics?


____________________________________________
Thomas Nichols, PhD
Director, Modelling & Genetics
GlaxoSmithKline Clinical Imaging Centre

Senior Research Fellow
Oxford University FMRIB Centre ------=_Part_20113_7950987.1192007417532-- ========================================================================= Date: Wed, 10 Oct 2007 11:42:37 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matthew Webster <[log in to unmask]> Subject: Re: probtrackx GUI issue on OSX cluster In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, It looks like the problem is with the batch call - the "-q" option doesn't seem to be available for the osx batch command. If you remove "-q long.q" from the batch command in fdt.tcl, I think it should batch correctly. Matthew > hello list, > > sorry to keep bugging you all, but we are making a lot of progress > with our cluster, thanks to this list, and we are almost fully > functional which is very exciting, as we are seeing some very > impressive speeds on the larger jobs (especially bedpost!) > > i implemented the symbolic links which were suggested to me in > earlier posts, and now we can run bedpostx jobs and FEAT jobs fine > from the gui, however we seem to have hit a snag with probtrackx. > here is the information i got from the user in question, please > advise what the problem could be, as it works fine from the command > line (although for some reason this job gets run on the head node > only and never gets submitted through the qsub process to the > cluster.) > > Running probtrackx from the GUI creates the appropriate output > directory, but only contains fdt_log.tcl and fdt_script.sh > > If I open fdt_script.sh and paste only the probtrackx command into > the terminal command line, the job runs perfectly. This is the > command that I paste: > > /common/fsl/bin/probtrackx --mode=seedmask -x /Volumes/Clinical/ > homes/tyanagihara/sweat/DTI/mask-dump/seed-internalCapsule- > mask.hdr -l -c 0.2 -S 2000 --steplength=0.5 -P 5000 --stop=/ > Volumes/Clinical/homes/tyanagihara/sweat/DTI/mask-dump/target1- > mask.hdr --forcedir --opd -s > > This is the error I get when running from the GUI: > usage: at [-q x] [-f file] [-m] time > at -c job [job ...] > at [-f file] -t [[CC]YY]MMDDhhmm[.SS] > at -r job [job ...] > at -l -q queuename > at -l [job ...] > atq [-q x] [-v] > atrm job [job ...] > batch [-f file] [-m] > usage: at [-q x] [-f file] [-m] time > at -c job [job ...] > at [-f file] -t [[CC]YY]MMDDhhmm[.SS] > at -r job [job ...] > at -l -q queuename > at -l [job ...] > atq [-q x] [-v] > atrm job [job ...] > batch [-f file] [-m] > while executing > "exec batch -q long.q ${filebase}_script.sh" > ("probtrackx" arm line 144) > invoked from within > "switch -- $probtrack(tool) { > eddy_current { > global eddy > > set errorStr "" > if { $eddy(input) == "" } { set errorStr "You need to specify..." > (procedure "fdt:apply" line 5) > invoked from within > "fdt:apply .fdt keep" > invoked from within > ".fdt.apply invoke" > ("uplevel" body line 1) > invoked from within > "uplevel #0 [list $w invoke]" > (procedure "tk::ButtonUp" line 22) > invoked from within > "tk::ButtonUp .fdt.apply" > (command bound to event) ========================================================================= Date: Wed, 10 Oct 2007 12:04:59 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matthew Webster <[log in to unmask]> Subject: Re: Installation worked...then failed In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Could this be a DISPLAY problem - what is your DISPLAY enviromnent variable set to? > I installed FSL on my apple computer at the lab. Everything worked > very well until I removed the > laptop from the lab and the network at the lab. When I attempted > using it at home, it stopped > working... > > I get the following address: > > anne-sophie-champods-computer:~ debas$ fsl > Application initialization failed: couldn't connect to display > "localhost:11" > Error in startup script: couldn't connect to display "localhost:11" > while executing > "load /usr/local/fsl/extras/lib/tk8.4/../libtk8.4.dylib Tk" > ("package ifneeded" script) > invoked from within > "package require Tk" > invoked from within > "if { [ string match -nocase *wish* $MYSHELL ] } { > package require Tk > # tk_focusFollowsMouse > #bind Button { focus %W ; tk::ButtonEnter ..." > (file "/usr/local/fsl/tcl/fslstart.tcl" line 19) > invoked from within > "source [ file dirname [ info script ] ]/fslstart.tcl" > (file "/usr/local/fsl/tcl/fsl.tcl" line 71) > invoked from within > "source ${FSLDIR}/tcl/${origname}.tcl" > (file "/usr/local/fsl/bin/fsl" line 22) > > Hopefully someone could help. Thank you, > > Roman ========================================================================= Date: Wed, 10 Oct 2007 12:09:38 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matthew Webster <[log in to unmask]> Subject: Re: WARNING: posn not header size error In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit How did you convert the files to nifti? Can you upload an example NIFTI file (and the the original ANALYZE files) to http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi and let us know the 6 digit reference number? Thanks Matthew > regarding nifti problem posn error > Here's what I get: > > > fslstats net1 -r -R > ** WARNING: posn not header size (0, 348) > 0.464809 914.370178 -16512.000000 31457.000000 > > > > fslhd net1.nii.gz > ** WARNING: posn not header size (0, 348) > filename net1.nii.gz > > sizeof_hdr 348 > data_type INT16 > dim0 4 > dim1 64 > dim2 64 > dim3 28 > dim4 100 > dim5 1 > dim6 1 > dim7 1 > vox_units mm > time_units s > datatype 4 > nbyper 2 > bitpix 16 > pixdim0 0.0000000000 > pixdim1 3.1250000000 > pixdim2 3.1250000000 > pixdim3 3.9900000095 > pixdim4 1.0000000000 > pixdim5 0.0000000000 > pixdim6 0.0000000000 > pixdim7 0.0000000000 > vox_offset 352 > cal_max 1923.0000 > cal_min 0.0000 > scl_slope 1.000000 > scl_inter 0.000000 > phase_dim 0 > freq_dim 0 > slice_dim 0 > slice_name Unknown > slice_code 0 > slice_start 0 > slice_end 0 > slice_duration 0.000000 > time_offset 0.000000 > intent Unknown > intent_code 0 > intent_name > intent_p1 0.000000 > intent_p2 0.000000 > intent_p3 0.000000 > qform_name Aligned Anat > qform_code 2 > qto_xyz:1 -3.125000 0.000000 -0.000000 96.875000 > qto_xyz:2 0.000000 3.125000 -0.000000 -96.875000 > qto_xyz:3 0.000000 0.000000 3.990000 -51.869999 > qto_xyz:4 0.000000 0.000000 0.000000 1.000000 > qform_xorient Right-to-Left > qform_yorient Posterior-to-Anterior > qform_zorient Inferior-to-Superior > sform_name Aligned Anat > sform_code 2 > sto_xyz:1 -3.125000 0.000000 0.000000 96.875000 > sto_xyz:2 0.000000 3.125000 0.000000 -96.875000 > sto_xyz:3 0.000000 0.000000 3.990000 -51.869999 > sto_xyz:4 0.000000 0.000000 0.000000 1.000000 > sform_xorient Right-to-Left > sform_yorient Posterior-to-Anterior > sform_zorient Inferior-to-Superior > file_type NIFTI-1+ > file_code 1 > descrip FSL4.0 > aux_file > > If I do fslhd on the analyze version of the file, I get: > > fslhd net_func_axial_6 > > filename net_func_axial_6 > > sizeof_hdr 348 > data_type dsr > db_name net_func_axial_6 > extents 16384 > session_error 0 > regular r > hkey_un0 > dim0 4 > dim1 64 > dim2 64 > dim3 28 > dim4 100 > dim5 0 > dim6 0 > dim7 0 > vox_units mm > cal_units > unused1 0 > datatype 4 > bitpix 16 > pixdim0 4.0000000000 > pixdim1 3.1250000000 > pixdim2 3.1250000000 > pixdim3 3.9900000095 > pixdim4 1.0000000000 > pixdim5 0.0000000000 > pixdim6 0.0000000000 > pixdim7 0.0000000000 > vox_offset 0.0000 > funused1 1.0000 > funused2 0.0000 > funused3 0.0000 > cal_max 1963.0000 > cal_min 0.0000 > compressed 0 > verified 0 > glmax 1963 > glmin 0 > descrip Bookheimer^NET > aux_file > orient 0 > originator > origin1 0 > origin2 0 > origin3 0 > generated (X)MedCon > scannum 1 > patient_id net_s1_1 > exp_date > exp_time > hist_un0 > views 0 > vols_added 0 > start_field 0 > field_skip 0 > omax 0 > omin 0 > smin 0 > smin 0 > file_type ANALYZE-7.5 > file_code 0 > ========================================================================= Date: Wed, 10 Oct 2007 14:07:56 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Duncan Mortimer <[log in to unmask]> Subject: Re: Installing FSL 4.0 from uncompiled source In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, You need to look in the 'build.log' file to see what actually failed to build. Searching for 'Error' should quickly take you to the problem area. Extras includes many sub-projects so it could be many different things. Duncan On 26 Sep 2007, at 21:55, Morgan Wolfe wrote: > To Whom It Concerns: > > I am trying to install FSL 4.0 on a Linux machine and received the > following > errors when building. > > ./build > Building projects - see build.log file for progress... > Finished build : end of log file shows ... > > > > !!ERROR in BUILD!! > Could not make the following projects successfully: > extras fast > > I am not sure what to do to correct these errors. Any assistance > woudl be > appreciated -- Duncan A B Mortimer DPhil MChem Computing Officer, FMRIB Centre, University of Oxford, John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK. Tel: (0)1865 222713 WWW: http://www.fmrib.ox.ac.uk/~duncan email: [log in to unmask] ========================================================================= Date: Wed, 10 Oct 2007 14:10:00 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Duncan Mortimer <[log in to unmask]> Subject: Re: Error while compiling FSL @ SuSE 10.2 (Linux 2.6.18.8-0.5-bigsmp i686) In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit I don't see any attachment - could you look through the 'build.log' file searching for 'Error'. Could you then include 5 lines either side of the error message in your reply. Duncan On 26 Sep 2007, at 21:38, Arthur Pilacinski wrote: > Hi, > > > I cannot compile FSL at SuSE 10.2 (Linux 2.6.18.8-0.5-bigsmp > i686). I am > not-a-very-experienced Linux user, so I'm not certain how to > interprete the > installation log (attached to this email). I someone could help me > with > that, I'd be grateful. > > > Kind regards, > > Arthur Duncan Mortimer [log in to unmask] ========================================================================= Date: Wed, 10 Oct 2007 09:28:41 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Chris Petty <[log in to unmask]> Subject: Re: first_utils MIME-Version: 1.0 Content-Type: text/plain Makes sense, thanks. So would the following designs below get what I expect? A group test, and take 2 scores into account my design would look like this: Group EV1 score1 score2 1 1 10 50 1 1 20 90 1 -1 10 200 1 -1 2 5 1 1 4 5 1 -1 30 8 Contrast = 1 0 0 Or a test between 3 groups may be: Group EV 1 1 1 1 1 2 1 2 1 3 1 3 Contrast = 1 -----Original Message----- From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Brian Patenaude Sent: Wednesday, October 10, 2007 4:15 AM To: [log in to unmask] Subject: Re: [FSL] first_utils Hi, The restriction is that no linear combination of the columns sum to a column of ones. first_utils demeans the design matrix such that the first column of the design matrix is all ones. This is why splitting the groups into into multiple columns causes an error. first_utils will demean all regressors internally. Cheers, Brian > ok, the preliminary problem .. i needed to use 1 EV with 1 or -1 for > the group difference. > > apparently the group difference needed to be in 1 ev and not multiple > evs, but how would you add an additional regressor if you can only > use > 1 EV? ========================================================================= Date: Wed, 10 Oct 2007 09:33:12 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Roman Loonis <[log in to unmask]> Subject: Re: Installation worked...then failed In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_10678_16906801.1192023192259" ------=_Part_10678_16906801.1192023192259 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline I am not sure exactly what that means. But I found out, that when X11 is already running, FSL works fine. If not, however, the program is unable to open X11 on its own. How can I change that? On 10/10/07, Matthew Webster <[log in to unmask]> wrote: > > Could this be a DISPLAY problem - what is your DISPLAY enviromnent > variable set to? > > I installed FSL on my apple computer at the lab. Everything worked > > very well until I removed the > > laptop from the lab and the network at the lab. When I attempted > > using it at home, it stopped > > working... > > > > I get the following address: > > > > anne-sophie-champods-computer:~ debas$ fsl > > Application initialization failed: couldn't connect to display > > "localhost:11" > > Error in startup script: couldn't connect to display "localhost:11" > > while executing > > "load /usr/local/fsl/extras/lib/tk8.4/../libtk8.4.dylib Tk" > > ("package ifneeded" script) > > invoked from within > > "package require Tk" > > invoked from within > > "if { [ string match -nocase *wish* $MYSHELL ] } { > > package require Tk > > # tk_focusFollowsMouse > > #bind Button { focus %W ; tk::ButtonEnter ..." > > (file "/usr/local/fsl/tcl/fslstart.tcl" line 19) > > invoked from within > > "source [ file dirname [ info script ] ]/fslstart.tcl" > > (file "/usr/local/fsl/tcl/fsl.tcl" line 71) > > invoked from within > > "source ${FSLDIR}/tcl/${origname}.tcl" > > (file "/usr/local/fsl/bin/fsl" line 22) > > > > Hopefully someone could help. Thank you, > > > > Roman > ------=_Part_10678_16906801.1192023192259 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline I am not sure exactly what that means. But I found out, that when X11 is already running, FSL works fine. If not, however, the program is unable to open X11 on its own. How can I change that?

On 10/10/07, Matthew Webster <[log in to unmask]> wrote:
Could this be a DISPLAY problem - what is your DISPLAY enviromnent
variable set to?
> I installed FSL on my apple computer at the lab. Everything worked
> very well until I removed the
> laptop from the lab and the network at the lab. When I attempted
> using it at home, it stopped
> working...
>
> I get the following address:
>
>  anne-sophie-champods-computer:~ debas$ fsl
> Application initialization failed: couldn't connect to display
> "localhost:11"
> Error in startup script: couldn't connect to display "localhost:11"
>     while executing
> "load /usr/local/fsl/extras/lib/tk8.4/../libtk8.4.dylib Tk"
>     ("package ifneeded" script)
>     invoked from within
> "package require Tk"
>     invoked from within
> "if { [  string match -nocase *wish* $MYSHELL ] } {
>     package require Tk
> #    tk_focusFollowsMouse
> #bind Button <Enter> { focus %W ; tk::ButtonEnter ..."
>     (file "/usr/local/fsl/tcl/fslstart.tcl" line 19)
>     invoked from within
> "source [ file dirname [ info script ] ]/fslstart.tcl"
>     (file "/usr/local/fsl/tcl/fsl.tcl" line 71)
>     invoked from within
> "source ${FSLDIR}/tcl/${origname}.tcl"
>     (file "/usr/local/fsl/bin/fsl" line 22)
>
> Hopefully someone could help.  Thank you,
>
> Roman

------=_Part_10678_16906801.1192023192259-- ========================================================================= Date: Wed, 10 Oct 2007 15:39:51 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matthew Webster <[log in to unmask]> Subject: Re: Installation worked...then failed In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: multipart/alternative; boundary=Apple-Mail-4--1028373148 --Apple-Mail-4--1028373148 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Hi, On the mac, you can go to System Preferences/Accounts/Login Items/ and add X11.app to the list - X11 will now automatically launch when you login.. Hope this helps Matt On 10 Oct 2007, at 14:33, Roman Loonis wrote: > I am not sure exactly what that means. But I found out, that when > X11 is already running, FSL works fine. If not, however, the > program is unable to open X11 on its own. How can I change that? > > On 10/10/07, Matthew Webster <[log in to unmask]> wrote: > Could this be a DISPLAY problem - what is your DISPLAY enviromnent > variable set to? > > I installed FSL on my apple computer at the lab. Everything worked > > very well until I removed the > > laptop from the lab and the network at the lab. When I attempted > > using it at home, it stopped > > working... > > > > I get the following address: > > > > anne-sophie-champods-computer:~ debas$ fsl > > Application initialization failed: couldn't connect to display > > "localhost:11" > > Error in startup script: couldn't connect to display "localhost:11" > > while executing > > "load /usr/local/fsl/extras/lib/tk8.4/../libtk8.4.dylib Tk" > > ("package ifneeded" script) > > invoked from within > > "package require Tk" > > invoked from within > > "if { [ string match -nocase *wish* $MYSHELL ] } { > > package require Tk > > # tk_focusFollowsMouse > > #bind Button { focus %W ; tk::ButtonEnter ..." > > (file "/usr/local/fsl/tcl/fslstart.tcl" line 19) > > invoked from within > > "source [ file dirname [ info script ] ]/fslstart.tcl" > > (file "/usr/local/fsl/tcl/fsl.tcl" line 71) > > invoked from within > > "source ${FSLDIR}/tcl/${origname}.tcl" > > (file "/usr/local/fsl/bin/fsl" line 22) > > > > Hopefully someone could help. Thank you, > > > > Roman > --Apple-Mail-4--1028373148 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1
Hi,
=A0 =A0=A0 On = the mac, you can go to=A0
System Preferences/Accounts/Login = Items/
and add X11.app to the list - X11 will now = automatically launch when you login..

Hope this = helps

Matt
On 10 Oct = 2007, at 14:33, Roman Loonis wrote:

I am not = sure exactly what that means. But I found out, that when X11 is already = running, FSL works fine. If not, however, the program is unable to open = X11 on its own. How can I change that?

On 10/10/07, Matthew = Webster <[log in to unmask]> = wrote:
Could this = be a DISPLAY problem - what is your DISPLAY enviromnent
variable set = to?
> I installed FSL on my apple computer at the lab. Everything = worked
> very well until I removed the
> laptop from the = lab and the network at the lab. When I attempted
> using it at = home, it stopped
> working...
>
> I get the following = address:
>
>=A0=A0anne-sophie-champods-computer:~ debas$ = fsl
> Application initialization failed: couldn't connect to = display
> "localhost:11"
> Error in startup script: couldn't = connect to display "localhost:11"
>=A0=A0=A0=A0 while = executing
> "load = /usr/local/fsl/extras/lib/tk8.4/../libtk8.4.dylib Tk"
>=A0=A0=A0=A0 = ("package ifneeded" script)
>=A0=A0=A0=A0 invoked from = within
> "package require Tk"
>=A0=A0=A0=A0 invoked from = within
> "if { [=A0=A0string match -nocase *wish* $MYSHELL ] } = {
>=A0=A0=A0=A0 package require Tk
> = #=A0=A0=A0=A0tk_focusFollowsMouse
> #bind Button <Enter> { = focus %W ; tk::ButtonEnter ..."
>=A0=A0=A0=A0 (file = "/usr/local/fsl/tcl/fslstart.tcl" line 19)
>=A0=A0=A0=A0 invoked = from within
> "source [ file dirname [ info script ] = ]/fslstart.tcl"
>=A0=A0=A0=A0 (file "/usr/local/fsl/tcl/fsl.tcl" = line 71)
>=A0=A0=A0=A0 invoked from within
> "source = ${FSLDIR}/tcl/${origname}.tcl"
>=A0=A0=A0=A0 (file = "/usr/local/fsl/bin/fsl" line 22)
>
> Hopefully someone = could help.=A0=A0Thank you,
>
> = Roman


= --Apple-Mail-4--1028373148-- ========================================================================= Date: Wed, 10 Oct 2007 15:41:13 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Duncan Mortimer <[log in to unmask]> Subject: Re: FSL installation under Windows XP In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, The user 'fsl' in the Virtual machine doesn't have 'sudoer' privileges. You can add the fsl user to the sudoer's list, but this an advanced procedure, and shouldn't be carried out unless you are confident with using the 'vi' editor. My recommendation is to install as the root user as follows, At a terminal: su - (enter the root password) cd ~fsl/Desktop/ ./fsl_installer.sh (enter details of installation as requested) exit cd ~/Desktop ./fsl_installer.sh -e I've updated the installation instructions to reflect this procedure. Duncan On 4 Oct 2007, at 09:28, James Gibbon wrote: > On Thu, 4 Oct 2007 09:56:05 +0200 > Markus Gschwind <[log in to unmask]> wrote: > >> No I like to unpack/untar "fsl-4.0.1-centos5_32.tar.gz" (now on >> Desktop of the Centos5) by the means of the fls_installer.sh, >> but it tells me that: >> "fsl is not in the sudoers file" >> >> What went wrong? How can I fix that? > > Hi, > > I'm not familiar with the script, but it essentially means that > the fsl user does not have the necessary privilege to run as > root, by means of the 'sudo' facility. It looks like the script > assumes that it does. > > Your system administrator may be willing to add the fsl user to > the 'sudoers' config file to overcome this. > > James > > This message has been checked for viruses but the contents of an > attachment > may still contain software viruses, which could damage your > computer system: > you are advised to perform your own checks. Email communications > with the > University of Nottingham may be monitored as permitted by UK > legislation. -- Duncan A B Mortimer DPhil MChem Computing Officer, FMRIB Centre, University of Oxford, John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK. Tel: (0)1865 222713 WWW: http://www.fmrib.ox.ac.uk/~duncan email: [log in to unmask] ========================================================================= Date: Wed, 10 Oct 2007 17:07:27 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Markus Gschwind <[log in to unmask]> Subject: Re: FSL installation under Windows XP In-Reply-To: <[log in to unmask]> MIME-version: 1.0 Content-transfer-encoding: 7BIT Content-disposition: inline Content-type: text/plain; charset=ISO-8859-1; DelSp=Yes; format=flowed Hi! Thanks! But meanwhile I've got it already running ;-) I just like to mention that I also had some problems with "gcc". It is not installed in your precompiled version of Centos. But while installing FSL you have to ad it. Given the fact that the FSL-Windows users are normaly not so enormously familiar with Linux, I think it would be a great possibility either to mention that the gcc installation has to be done (and how) or to install it directly in the precompiled version of Centos. Thanks again! Markus Quoting Duncan Mortimer <[log in to unmask]>: > Hi, > > The user 'fsl' in the Virtual machine doesn't have 'sudoer' privileges. > You can add the fsl user to the sudoer's list, but this an advanced > procedure, and shouldn't be carried out unless you are confident with > using the 'vi' editor. > My recommendation is to install as the root user as follows, > At a terminal: > > su - > (enter the root password) > cd ~fsl/Desktop/ > ./fsl_installer.sh > (enter details of installation as requested) > exit > cd ~/Desktop > ./fsl_installer.sh -e > > I've updated the installation instructions to reflect this procedure. > > Duncan > > > On 4 Oct 2007, at 09:28, James Gibbon wrote: > >> On Thu, 4 Oct 2007 09:56:05 +0200 >> Markus Gschwind <[log in to unmask]> wrote: >> >>> No I like to unpack/untar "fsl-4.0.1-centos5_32.tar.gz" (now on >>> Desktop of the Centos5) by the means of the fls_installer.sh, >>> but it tells me that: >>> "fsl is not in the sudoers file" >>> >>> What went wrong? How can I fix that? >> >> Hi, >> >> I'm not familiar with the script, but it essentially means that >> the fsl user does not have the necessary privilege to run as >> root, by means of the 'sudo' facility. It looks like the script >> assumes that it does. >> >> Your system administrator may be willing to add the fsl user to >> the 'sudoers' config file to overcome this. >> >> James >> >> This message has been checked for viruses but the contents of an attachment >> may still contain software viruses, which could damage your computer system: >> you are advised to perform your own checks. Email communications with the >> University of Nottingham may be monitored as permitted by UK legislation. > > -- > Duncan A B Mortimer DPhil MChem > Computing Officer, FMRIB Centre, University of Oxford, > John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK. > Tel: (0)1865 222713 > WWW: http://www.fmrib.ox.ac.uk/~duncan email: [log in to unmask] -- Markus Gschwind, M.D. Laboratory for Neurology and Imaging of Cognition Dept of Neurosciences University Medical Center (CMU) 1 Michel-Servet - 1211 GENEVA - CH Tel 0041 (0) 22 379 5324 Fax 0041 (0) 22 379 5402 email: [log in to unmask] http://labnic.unige.ch ========================================================================= Date: Wed, 10 Oct 2007 16:24:06 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: empty EVs in lower level analysis In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - yes, this should work - it's just a case of persuading FEAT that all is well. We've just made some minor amendments to the scripts to allow empty EVs to run through ok and even contrasts involving empty EVs (effectively assumes that the relevant PE is zero even though this can't be estimated properly). This will be released in the next patch release, probably next week. Cheers, Steve. On 28 Sep 2007, at 19:30, Eric Claus wrote: > FSLers, > I am trying to run first level analyses for each run of a 3-run > fMRI study. > Within each run, there are several trial types, but because the > classification of a trial depends on the participant's response, > there are > some cases where 1 or more trial types do not exist in a given > run. For > each trial type, I have a separate EV file and for the cases in > which no > trials exist, the empty EV option in FSL is chosen. > > 2 questions: > > 1. When trying to run the subject's model, I get a rank deficiency > message > which seems to be expected given the empty EVs. If all the other > EVs in the > model look correct (i.e. are not linear combinations of one > another), are > the results of the analysis valid? > > 2. If I have 4 EVs: EV1 EV2 EV3 EV4 and a given subject has an > empty EV for > EV2, how does this affect any contrast with EV2? For example, if a > contrast > of 1 -1 0 0 is used, is the resulting statistical map equivalent to > the > mean/fit for EV1 or should the stat map be zero, given that the two > conditions are not really being compared? Further, if the > participant has > valid trials for runs 2 and 3 in both EV1 and EV2, when I do the fixed > effects analysis for this subject, is the run1 contrast included in > the mean? > > Thanks in advance, > Eric ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Wed, 10 Oct 2007 16:36:17 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Carlos Faraco <[log in to unmask]> Subject: DICOM converter Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" I having been to trying to find a good GE DICOM converter (either to NIFT= I or analyze is fine). Our in-house MATLAB program, which we use for DICOM = to analyze conversion for SPM2 & 5 does not seem to work well with FSL, = A and P are switeched when images are displayed. I guess this may have to do something with FSL ignoring certain information in the file, as mentioned= in the mricron website. DCM2NII converter from mricron works well for the anatomicals, but not for one of our current studies which has 6600 functional DICOMs. There is some limitation within linux that does not al= low an operation to be performed on so many files (i.e, argument list too lon= g). I have also tried LONI, but this does not seem to work. Does anybody have= any suggestions? I have been trying to re-process some data through FEAT,= but have been reluctant to do so because of the A to P issue mentioned above. Maybe I could just re-orient the images, but I imagine there would= be some problems with this since the coordinate information within the file = is not what is being shown. Thanks, Carlos ========================================================================= Date: Wed, 10 Oct 2007 16:35:12 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Dave Flitney <[log in to unmask]> Subject: Re: fslview: segmentation fault In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: multipart/alternative; boundary=Apple-Mail-10--1025052325 --Apple-Mail-10--1025052325 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Naj, I've switched maintenance efforts to Version 3.0.x now and, unless you have a really important reason for not using the new version, I'd ask that you upgrade FSLView. Debian packages are available, see http://apsy.gse.uni-magdeburg.de/fsl for details. On 8 Oct 2007, at 18:06, Najmeh Khalili M. wrote: > Hi Dave, > > I have the fslview (Version 2.4pre0), running on: > Debian GNU/Linux 4.0 \n \l > > rosaline:~# uname -a > Linux rosaline 2.6.22.2-i686-smp #1 SMP Tue Sep 25 14:47:07 EDT > 2007 i686 GNU/Linux > > Thanks, > Naj > > > > On Mon, 8 Oct 2007, Dave Flitney wrote: > >> Naj, >> >> It opens fine on the platforms I've tried so far - FSLView 3.0 on: >> MacOSX (PPC and Intel); Fedora Core 5 (64bit); and Centos4 (32bit). >> >> Which OS & FSLView versions are you using? >> >> On 5 Oct 2007, at 16:51, Najmeh Khalili M. wrote: >> >>> Hi Dave, >>> >>> Ref number is: >>> 943835 >>> >>> This is supposed to be a mask. It is not empty at fslstats >>> indicates non-zero content. >> >> Didn't think it was. That was a different poster :-) >> >>> Thanks >>> Naj >> >> -- >> Cheers, Dave >> >> Dave Flitney, IT Manager >> Oxford Centre for Functional MRI of the Brain >> E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 >> URL: http://www.fmrib.ox.ac.uk/~flitney >> >> >> -- Cheers, Dave Dave Flitney, IT Manager Oxford Centre for Functional MRI of the Brain E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 URL: http://www.fmrib.ox.ac.uk/~flitney --Apple-Mail-10--1025052325 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1 Naj,

I've switched maintenance = efforts to Version 3.0.x now and, unless you have a really important = reason for not using the new version, I'd ask that you upgrade = FSLView.=A0Debian packages are available, see=A0http://apsy.gse.uni-magdebur= g.de/fsl for details.

On 8 Oct 2007, at = 18:06, Najmeh Khalili M. wrote:

Hi Dave,

I have the=A0 fslview (Version 2.4pre0), = running on:
Debian GNU/Linux 4.0 \n = \l

rosaline:~# uname -a
Linux = rosaline 2.6.22.2-i686-smp #1 SMP Tue Sep 25 14:47:07 EDT
2007 i686 GNU/Linux

Naj



On Mon, 8 Oct 2007, Dave Flitney wrote:

Naj,

It opens = fine on the platforms I've tried so far - FSLView 3.0 on:
MacOSX (PPC and Intel); Fedora Core 5 (64bit); and = Centos4 (32bit).

Which OS & FSLView versions are you = using?

On 5 Oct 2007, at 16:51, Najmeh Khalili M. = wrote:

=
Hi Dave,

Ref = number is:
943835

This is = supposed to be a mask. It is not empty at fslstats
indicates non-zero content.

Didn't = think it was. That was a different poster :-)

Thanks
Naj

--
Cheers, Dave

Dave Flitney, IT = Manager
Oxford Centre for Functional MRI = of the Brain
E:[log in to unmask] = W:+44-1865-222713 F:+44-1865-222717




=
Dave Flitney, = IT Manager
Oxford Centre for Functional MRI = of the Brain
E:[log in to unmask] = W:+44-1865-222713 F:+44-1865-222717

=

= --Apple-Mail-10--1025052325-- ========================================================================= Date: Wed, 10 Oct 2007 17:43:45 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Sebastian Moeller <[log in to unmask]> Subject: Re: DICOM converter In-Reply-To: <[log in to unmask]> MIME-version: 1.0 Content-type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-transfer-encoding: 7BIT Hi there, not an expert at all, but... First FreeSurfer is really great in handling dicoms (even I only tried siemens ones), and its conversion tool mri_convert is a joy to work with. On Oct 10, 2007, at 5:36 PM, Carlos Faraco wrote: > I having been to trying to find a good GE DICOM converter (either > to NIFTI > or analyze is fine). Our in-house MATLAB program, which we use for > DICOM to > analyze conversion for SPM2 & 5 does not seem to work well with > FSL, A and P > are switeched when images are displayed. I guess this may have to do > something with FSL ignoring certain information in the file, as > mentioned in > the mricron website. DCM2NII converter from mricron works well for the > anatomicals, but not for one of our current studies which has 6600 > functional DICOMs. There is some limitation within linux that does > not allow > an operation to be performed on so many files (i.e, argument list > too long). You are lucky, that one is severly improved with the just released kernel 2.6.23, so download, compile an stick to the tools you already know, alternatively I seem to recall that xargs supposedly can be used to work around the argument list problem (no first hand experience though). > I have also tried LONI, but this does not seem to work. Does > anybody have > any suggestions? I have been trying to re-process some data through > FEAT, > but have been reluctant to do so because of the A to P issue mentioned > above. Maybe I could just re-orient the images, but I imagine there > would be > some problems with this since the coordinate information within the > file is > not what is being shown. > > Thanks, > > Carlos -- Sebastian Moeller Tel.: 04 21 - 2 18 - 78 38 oder 96 91 Fax.: 04 21 - 2 18 - 90 04 GSM: 0 15 77 - 1 90 31 41 [log in to unmask] AG Kreiter / FB 2 Institut fuer Hirnforschung III Abteilung Theoretische Neurobiologie Universitaet Bremen Biogarten Hochschulring 16a Postfach 33 04 40 28359 Bremen ========================================================================= Date: Wed, 10 Oct 2007 16:42:28 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Duncan Mortimer <[log in to unmask]> Subject: Re: FSL installation under Windows XP In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, GCC isn't actually necessary, the message you see on the screen during the sourcing of the fsl.sh file is just a warning and can safely be ignored. However, we have fixed the script in the current development version, to prevent this warning. The next release should include this change. As for installing GCC, you can use the 'Add/Remove Software' GUI to add the 'Development Tools'. They aren't included in the VM for reasons of space, as they would add significantly to the size of the download. Duncan On 10 Oct 2007, at 16:07, Markus Gschwind wrote: > Hi! > Thanks! But meanwhile I've got it already running ;-) > > I just like to mention that I also had some problems with "gcc". > It is not installed in your precompiled version of Centos. > But while installing FSL you have to ad it. > > Given the fact that the FSL-Windows users are normaly not so > enormously familiar with Linux, I think it would be a great > possibility either to mention that the gcc installation has to be > done (and how) or to install it directly in the precompiled version > of Centos. > > Thanks again! > Markus > > > Quoting Duncan Mortimer <[log in to unmask]>: > >> Hi, >> >> The user 'fsl' in the Virtual machine doesn't have 'sudoer' >> privileges. >> You can add the fsl user to the sudoer's list, but this an advanced >> procedure, and shouldn't be carried out unless you are confident with >> using the 'vi' editor. >> My recommendation is to install as the root user as follows, >> At a terminal: >> >> su - >> (enter the root password) >> cd ~fsl/Desktop/ >> ./fsl_installer.sh >> (enter details of installation as requested) >> exit >> cd ~/Desktop >> ./fsl_installer.sh -e >> >> I've updated the installation instructions to reflect this procedure. >> >> Duncan >> >> >> On 4 Oct 2007, at 09:28, James Gibbon wrote: >> >>> On Thu, 4 Oct 2007 09:56:05 +0200 >>> Markus Gschwind <[log in to unmask]> wrote: >>> >>>> No I like to unpack/untar "fsl-4.0.1-centos5_32.tar.gz" (now on >>>> Desktop of the Centos5) by the means of the fls_installer.sh, >>>> but it tells me that: >>>> "fsl is not in the sudoers file" >>>> >>>> What went wrong? How can I fix that? >>> >>> Hi, >>> >>> I'm not familiar with the script, but it essentially means that >>> the fsl user does not have the necessary privilege to run as >>> root, by means of the 'sudo' facility. It looks like the script >>> assumes that it does. >>> >>> Your system administrator may be willing to add the fsl user to >>> the 'sudoers' config file to overcome this. >>> >>> James >>> >>> This message has been checked for viruses but the contents of an >>> attachment >>> may still contain software viruses, which could damage your >>> computer system: >>> you are advised to perform your own checks. Email communications >>> with the >>> University of Nottingham may be monitored as permitted by UK >>> legislation. >> >> -- >> Duncan A B Mortimer DPhil MChem >> Computing Officer, FMRIB Centre, University of >> Oxford, >> John Radcliffe Hospital, Headington, Oxford OX3 9DU, >> UK. >> Tel: (0)1865 222713 >> WWW: http://www.fmrib.ox.ac.uk/~duncan email: >> [log in to unmask] > > > > -- > > Markus Gschwind, M.D. > Laboratory for Neurology and Imaging of Cognition > Dept of Neurosciences > University Medical Center (CMU) > 1 Michel-Servet - 1211 GENEVA - CH > > Tel 0041 (0) 22 379 5324 > Fax 0041 (0) 22 379 5402 > email: [log in to unmask] > http://labnic.unige.ch -- Duncan A B Mortimer DPhil MChem Computing Officer, FMRIB Centre, University of Oxford, John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK. Tel: (0)1865 222713 WWW: http://www.fmrib.ox.ac.uk/~duncan email: [log in to unmask] ========================================================================= Date: Wed, 10 Oct 2007 17:43:54 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Michael Hanke <[log in to unmask]> Subject: Re: DICOM converter In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Hi, On Wed, Oct 10, 2007 at 04:36:17PM +0100, Carlos Faraco wrote: > I having been to trying to find a good GE DICOM converter (either to NIFTI > or analyze is fine). Our in-house MATLAB program, which we use for DICOM to > analyze conversion for SPM2 & 5 does not seem to work well with FSL, A and P > are switeched when images are displayed. I guess this may have to do > something with FSL ignoring certain information in the file, as mentioned in > the mricron website. DCM2NII converter from mricron works well for the > anatomicals, but not for one of our current studies which has 6600 > functional DICOMs. There is some limitation within linux that does not allow > an operation to be performed on so many files (i.e, argument list too long). > I have also tried LONI, but this does not seem to work. Does anybody have > any suggestions? I have been trying to re-process some data through FEAT, > but have been reluctant to do so because of the A to P issue mentioned > above. Maybe I could just re-orient the images, but I imagine there would be > some problems with this since the coordinate information within the file is > not what is being shown. You could try dicomnifti: http://cbi.nyu.edu/software/dinifti.php You can simply specify the directory containing your dicoms, so there should be no problems concerning the length of the argument list. I have no experience of using it with GE dicoms, though. Cheers, Michael -- GPG key: 1024D/3144BE0F Michael Hanke http://apsy.gse.uni-magdeburg.de/hanke ICQ: 48230050 ========================================================================= Date: Wed, 10 Oct 2007 11:52:01 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Hedok Lee <[log in to unmask]> Subject: Re: DICOM converter In-Reply-To: <[log in to unmask]> MIME-version: 1.0 Content-type: text/plain; charset=ISO-8859-1; format=flowed Content-transfer-encoding: 7BIT Dear Carlos: I've been using SPM5 to convert DICOM(GE signa scanner) to nifit, and it has been working well. I mainly use FSL for anatomical and DTI analyses so I can't be certain about your functional images. I remember when I used LONI debabeler, I had a similar problem with A-P being reversed. You can also try "MRIConvert" http://lcni.uoregon.edu/~jolinda/MRIConvert/ or Freesurfer's MRI convert works well too. http://surfer.nmr.mgh.harvard.edu/ Hope this helps, Hedok Carlos Faraco wrote: > I having been to trying to find a good GE DICOM converter (either to NIFTI > or analyze is fine). Our in-house MATLAB program, which we use for DICOM to > analyze conversion for SPM2 & 5 does not seem to work well with FSL, A and P > are switeched when images are displayed. I guess this may have to do > something with FSL ignoring certain information in the file, as mentioned in > the mricron website. DCM2NII converter from mricron works well for the > anatomicals, but not for one of our current studies which has 6600 > functional DICOMs. There is some limitation within linux that does not allow > an operation to be performed on so many files (i.e, argument list too long). > I have also tried LONI, but this does not seem to work. Does anybody have > any suggestions? I have been trying to re-process some data through FEAT, > but have been reluctant to do so because of the A to P issue mentioned > above. Maybe I could just re-orient the images, but I imagine there would be > some problems with this since the coordinate information within the file is > not what is being shown. > > Thanks, > > Carlos > ========================================================================= Date: Wed, 10 Oct 2007 16:47:57 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Dave Flitney <[log in to unmask]> Subject: Re: Running TBSS on four CPUs In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: multipart/alternative; boundary=Apple-Mail-11--1024287471 --Apple-Mail-11--1024287471 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Thomas, The tbss_2_reg script creates a file full of the commands to be run then uses fsl_sub to, either, run all the commands serially or submit them in parallel to your SGE compute cluster - step 1 is to remove any old ".commands" file to avoid confusion :-) hence the message. One crude way to use all your CPUs would be to comment out the fsl_sub command then manually run the commands in the ".command" file N-at-a-time. On 8 Oct 2007, at 19:40, Thomas Doring wrote: > Hi, > i am trying to run the registration step, tbss_2_reg -T, with all my 4 > CPUs but when i open the second terminal i just get this result: > [thomas@localhost TBSS]# tbss_2_reg -T > wc: .commands: No such file or directory > 16862 > > Somebody has an idea? > Thanks, > Thomas -- Cheers, Dave Dave Flitney, IT Manager Oxford Centre for Functional MRI of the Brain E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 URL: http://www.fmrib.ox.ac.uk/~flitney --Apple-Mail-11--1024287471 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1 Thomas,

The tbss_2_reg script = creates a file full of the commands to be run then uses fsl_sub to, = either, run all the commands serially or submit them in parallel to your = SGE compute cluster - step 1 is to remove any old ".commands" file to = avoid confusion :-) hence the message. One crude way to use all your = CPUs would be to comment out the fsl_sub command then manually run the = commands in the ".command" file N-at-a-time.

On 8 Oct 2007, at 19:40, = Thomas Doring wrote:

Hi,=A0
i am = trying to run the registration step, tbss_2_reg -T, with all my = 4
CPUs but when i open the second terminal i just = get this result:
[thomas@localhost TBSS]# = tbss_2_reg -T
wc: .commands: No such file or = directory
16862

Somebody = has an idea?=A0
Thanks,
Thomas
=

Dave Flitney, = IT Manager
Oxford Centre for Functional MRI = of the Brain
E:[log in to unmask] = W:+44-1865-222713 F:+44-1865-222717

=

= --Apple-Mail-11--1024287471-- ========================================================================= Date: Wed, 10 Oct 2007 13:52:45 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Najmeh Khalili M." <[log in to unmask]> Subject: Re: fslview: segmentation fault In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi Dave, It seems the debian package you pointed to doesn't have the recent fslview you suggested. Our sys admin is asking if you plan to make a more recent or the most recent version of flsview for the Debian "etch" distribution. We've tested the current available version of fslview for etch, first by installing the binary distribution and then by installing/building from source using apt-src. Thanks, Naj On Wed, 10 Oct 2007, Dave Flitney wrote: > Naj, > > I've switched maintenance efforts to Version 3.0.x now and, unless > you have a really important reason for not using the new version, I'd > ask that you upgrade FSLView. Debian packages are available, see > http://apsy.gse.uni-magdeburg.de/fsl for details. > > On 8 Oct 2007, at 18:06, Najmeh Khalili M. wrote: > > > Hi Dave, > > > > I have the fslview (Version 2.4pre0), running on: > > Debian GNU/Linux 4.0 \n \l > > > > rosaline:~# uname -a > > Linux rosaline 2.6.22.2-i686-smp #1 SMP Tue Sep 25 14:47:07 EDT > > 2007 i686 GNU/Linux > > > > Thanks, > > Naj > > > > > > > > On Mon, 8 Oct 2007, Dave Flitney wrote: > > > >> Naj, > >> > >> It opens fine on the platforms I've tried so far - FSLView 3.0 on: > >> MacOSX (PPC and Intel); Fedora Core 5 (64bit); and Centos4 (32bit). > >> > >> Which OS & FSLView versions are you using? > >> > >> On 5 Oct 2007, at 16:51, Najmeh Khalili M. wrote: > >> > >>> Hi Dave, > >>> > >>> Ref number is: > >>> 943835 > >>> > >>> This is supposed to be a mask. It is not empty at fslstats > >>> indicates non-zero content. > >> > >> Didn't think it was. That was a different poster :-) > >> > >>> Thanks > >>> Naj > >> > >> -- > >> Cheers, Dave > >> > >> Dave Flitney, IT Manager > >> Oxford Centre for Functional MRI of the Brain > >> E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 > >> URL: http://www.fmrib.ox.ac.uk/~flitney > >> > >> > >> > > -- > Cheers, Dave > > Dave Flitney, IT Manager > Oxford Centre for Functional MRI of the Brain > E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 > URL: http://www.fmrib.ox.ac.uk/~flitney > > > ========================================================================= Date: Wed, 10 Oct 2007 19:13:31 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Vishwadeep Ahluwalia <[log in to unmask]> Subject: Re: ICA - rescaling x-axis of FFT of TC Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Hi, I have 202 volumes of spiral in/out data with TR=3D2s. i ran melodic on t= his data. According to the previous mail in this topic, the highest freq that= can be represented corresponds to 101 cycles. Ive 8 blocks i.e 8 cycles, = so the 8th value in the text file should have a large value for the component corresponding to my expt., which i do get in many components. However in atleast 2 components( this is true in other= datasets as well) i get a peak at 0.15hz(~60cycles in 404s). Is th= is the peak corresponding to respiration? Also, since my sampling frequency is 0.5Hz(corresponds to 202 cycles in 404s), and the cardiac pulsation being much more than that, will i see the peak a= t the aliased frequency or will it not be shown at all? Ive uploaded the times series ,power spectrum and the ICmap of the 0.15Hz.The upload id is= 749962 Thanks -Vish ========================================================================= Date: Wed, 10 Oct 2007 19:20:03 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Carlos Faraco <[log in to unmask]> Subject: Re: DICOM converter Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" I have actually tried the DICOM converter from SPM5 and it says it cannot= determine what order the images is in and will just guess. Thank you, and= everyone else, for the suggestions. I will try them. -Carlos ========================================================================= Date: Wed, 10 Oct 2007 14:36:55 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Lokke Highstein <[log in to unmask]> Subject: Re: probtrackx GUI issue on OSX cluster In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit hi matthew, we don't have a long.q (or a short.q, veryshort.q, or verylong.q) we only are currently using the all.q so i went and changed all the references to "long.q" in the fdt.tcl file and edited them to all.q instead. i think that is what you are getting at. if you were suggesting that i remove any references to long.q by commenting out the whole line, i can do that but it seems like it would guarantee that the job wouldn't get submitted to the cluster then. unfortunately changing that didn't help with the GUI error, and even under command line use, the job still just ran on the head node (didn't get submitted to the SGE) here's the error output from the GUI: usage: at [-q x] [-f file] [-m] time at -c job [job ...] at [-f file] -t [[CC]YY]MMDDhhmm[.SS] at -r job [job ...] at -l -q queuename at -l [job ...] atq [-q x] [-v] atrm job [job ...] batch [-f file] [-m] usage: at [-q x] [-f file] [-m] time at -c job [job ...] at [-f file] -t [[CC]YY]MMDDhhmm[.SS] at -r job [job ...] at -l -q queuename at -l [job ...] atq [-q x] [-v] atrm job [job ...] batch [-f file] [-m] while executing "exec batch -q all.q ${filebase}_script.sh" ("probtrackx" arm line 144) invoked from within "switch -- $probtrack(tool) { eddy_current { global eddy set errorStr "" if { $eddy(input) == "" } { set errorStr "You need to specify..." (procedure "fdt:apply" line 5) invoked from within "fdt:apply .fdt keep" invoked from within ".fdt.apply invoke" ("uplevel" body line 1) invoked from within "uplevel #0 [list $w invoke]" (procedure "tk::ButtonUp" line 22) invoked from within "tk::ButtonUp .fdt.apply" (command bound to event) On Oct 10, 2007, at 6:42 AM, Matthew Webster wrote: > Hi, > It looks like the problem is with the batch call - the "-q" > option doesn't seem to be available for the osx batch command. If > you remove "-q long.q" from the batch command in fdt.tcl, I think > it should batch correctly. > > Matthew >> hello list, >> >> sorry to keep bugging you all, but we are making a lot of progress >> with our cluster, thanks to this list, and we are almost fully >> functional which is very exciting, as we are seeing some very >> impressive speeds on the larger jobs (especially bedpost!) >> >> i implemented the symbolic links which were suggested to me in >> earlier posts, and now we can run bedpostx jobs and FEAT jobs fine >> from the gui, however we seem to have hit a snag with probtrackx. >> here is the information i got from the user in question, please >> advise what the problem could be, as it works fine from the >> command line (although for some reason this job gets run on the >> head node only and never gets submitted through the qsub process >> to the cluster.) >> >> Running probtrackx from the GUI creates the appropriate output >> directory, but only contains fdt_log.tcl and fdt_script.sh >> >> If I open fdt_script.sh and paste only the probtrackx command into >> the terminal command line, the job runs perfectly. This is the >> command that I paste: >> >> /common/fsl/bin/probtrackx --mode=seedmask -x /Volumes/Clinical/ >> homes/tyanagihara/sweat/DTI/mask-dump/seed-internalCapsule- >> mask.hdr -l -c 0.2 -S 2000 --steplength=0.5 -P 5000 --stop=/ >> Volumes/Clinical/homes/tyanagihara/sweat/DTI/mask-dump/target1- >> mask.hdr --forcedir --opd -s >> >> This is the error I get when running from the GUI: >> usage: at [-q x] [-f file] [-m] time >> at -c job [job ...] >> at [-f file] -t [[CC]YY]MMDDhhmm[.SS] >> at -r job [job ...] >> at -l -q queuename >> at -l [job ...] >> atq [-q x] [-v] >> atrm job [job ...] >> batch [-f file] [-m] >> usage: at [-q x] [-f file] [-m] time >> at -c job [job ...] >> at [-f file] -t [[CC]YY]MMDDhhmm[.SS] >> at -r job [job ...] >> at -l -q queuename >> at -l [job ...] >> atq [-q x] [-v] >> atrm job [job ...] >> batch [-f file] [-m] >> while executing >> "exec batch -q long.q ${filebase}_script.sh" >> ("probtrackx" arm line 144) >> invoked from within >> "switch -- $probtrack(tool) { >> eddy_current { >> global eddy >> >> set errorStr "" >> if { $eddy(input) == "" } { set errorStr "You need to >> specify..." >> (procedure "fdt:apply" line 5) >> invoked from within >> "fdt:apply .fdt keep" >> invoked from within >> ".fdt.apply invoke" >> ("uplevel" body line 1) >> invoked from within >> "uplevel #0 [list $w invoke]" >> (procedure "tk::ButtonUp" line 22) >> invoked from within >> "tk::ButtonUp .fdt.apply" >> (command bound to event) ========================================================================= Date: Wed, 10 Oct 2007 11:49:42 -0700 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Dianne Patterson <[log in to unmask]> Subject: Re: DICOM converter In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_58456_16967042.1192042182659" ------=_Part_58456_16967042.1192042182659 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline We have GE dicom images and MRIconvert (mentioned before) is a godsend...fast, convenient, accurate.. I've only been using the windows version, but I'm very impressed. http://lcni.uoregon.edu/~jolinda/MRIConvert/ -Dianne On 10/10/07, Carlos Faraco <[log in to unmask]> wrote: > > I have actually tried the DICOM converter from SPM5 and it says it cannot > determine what order the images is in and will just guess. Thank you, and > everyone else, for the suggestions. I will try them. > > -Carlos > > -- Dianne Patterson, Ph.D. [log in to unmask] ERP Lab University of Arizona 621-3256 (Office) ------=_Part_58456_16967042.1192042182659 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline We have GE dicom images and MRIconvert (mentioned before) is a godsend...fast, convenient, accurate..
I've only been using the windows version, but I'm very impressed.
http://lcni.uoregon.edu/~jolinda/MRIConvert/

-Dianne

On 10/10/07, Carlos Faraco <[log in to unmask]> wrote:
I have actually tried the DICOM converter from SPM5 and it says it cannot
determine what order the images is in and will just guess. Thank you, and
everyone else, for the suggestions. I will try them.

-Carlos




--
Dianne Patterson, Ph.D.
[log in to unmask]
ERP Lab
University of Arizona
621-3256 (Office) ------=_Part_58456_16967042.1192042182659-- ========================================================================= Date: Wed, 10 Oct 2007 14:57:13 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Merry Mani <[log in to unmask]> Subject: Re: probtrackx GUI issue on OSX cluster In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, I had the same problem on OS X too. I showed it to my system administrator and he modified the fdt.tcl by changing the "exec batch -q all.q ${filebase}_script.sh" to "exec qsub -q all.q ${filebase} _script.sh". Now it works fine for me. -Merry On Oct 10, 2007, at 2:36 PM, Lokke Highstein wrote: > hi matthew, > > we don't have a long.q (or a short.q, veryshort.q, or verylong.q) > we only are currently using the all.q > > so i went and changed all the references to "long.q" in the fdt.tcl > file and edited them to all.q instead. i think that is what you > are getting at. if you were suggesting that i remove any > references to long.q by commenting out the whole line, i can do > that but it seems like it would guarantee that the job wouldn't get > submitted to the cluster then. > > unfortunately changing that didn't help with the GUI error, and > even under command line use, the job still just ran on the head > node (didn't get submitted to the SGE) > > here's the error output from the GUI: > > usage: at [-q x] [-f file] [-m] time > at -c job [job ...] > at [-f file] -t [[CC]YY]MMDDhhmm[.SS] > at -r job [job ...] > at -l -q queuename > at -l [job ...] > atq [-q x] [-v] > atrm job [job ...] > batch [-f file] [-m] > usage: at [-q x] [-f file] [-m] time > at -c job [job ...] > at [-f file] -t [[CC]YY]MMDDhhmm[.SS] > at -r job [job ...] > at -l -q queuename > at -l [job ...] > atq [-q x] [-v] > atrm job [job ...] > batch [-f file] [-m] > while executing > "exec batch -q all.q ${filebase}_script.sh" > ("probtrackx" arm line 144) > invoked from within > "switch -- $probtrack(tool) { > eddy_current { > global eddy > > set errorStr "" > if { $eddy(input) == "" } { set errorStr "You need to specify..." > (procedure "fdt:apply" line 5) > invoked from within > "fdt:apply .fdt keep" > invoked from within > ".fdt.apply invoke" > ("uplevel" body line 1) > invoked from within > "uplevel #0 [list $w invoke]" > (procedure "tk::ButtonUp" line 22) > invoked from within > "tk::ButtonUp .fdt.apply" > (command bound to event) > > On Oct 10, 2007, at 6:42 AM, Matthew Webster wrote: > >> Hi, >> It looks like the problem is with the batch call - the "-q" >> option doesn't seem to be available for the osx batch command. If >> you remove "-q long.q" from the batch command in fdt.tcl, I think >> it should batch correctly. >> >> Matthew >>> hello list, >>> >>> sorry to keep bugging you all, but we are making a lot of >>> progress with our cluster, thanks to this list, and we are almost >>> fully functional which is very exciting, as we are seeing some >>> very impressive speeds on the larger jobs (especially bedpost!) >>> >>> i implemented the symbolic links which were suggested to me in >>> earlier posts, and now we can run bedpostx jobs and FEAT jobs >>> fine from the gui, however we seem to have hit a snag with >>> probtrackx. here is the information i got from the user in >>> question, please advise what the problem could be, as it works >>> fine from the command line (although for some reason this job >>> gets run on the head node only and never gets submitted through >>> the qsub process to the cluster.) >>> >>> Running probtrackx from the GUI creates the appropriate output >>> directory, but only contains fdt_log.tcl and fdt_script.sh >>> >>> If I open fdt_script.sh and paste only the probtrackx command >>> into the terminal command line, the job runs perfectly. This is >>> the command that I paste: >>> >>> /common/fsl/bin/probtrackx --mode=seedmask -x /Volumes/Clinical/ >>> homes/tyanagihara/sweat/DTI/mask-dump/seed-internalCapsule- >>> mask.hdr -l -c 0.2 -S 2000 --steplength=0.5 -P 5000 --stop=/ >>> Volumes/Clinical/homes/tyanagihara/sweat/DTI/mask-dump/target1- >>> mask.hdr --forcedir --opd -s >>> >>> This is the error I get when running from the GUI: >>> usage: at [-q x] [-f file] [-m] time >>> at -c job [job ...] >>> at [-f file] -t [[CC]YY]MMDDhhmm[.SS] >>> at -r job [job ...] >>> at -l -q queuename >>> at -l [job ...] >>> atq [-q x] [-v] >>> atrm job [job ...] >>> batch [-f file] [-m] >>> usage: at [-q x] [-f file] [-m] time >>> at -c job [job ...] >>> at [-f file] -t [[CC]YY]MMDDhhmm[.SS] >>> at -r job [job ...] >>> at -l -q queuename >>> at -l [job ...] >>> atq [-q x] [-v] >>> atrm job [job ...] >>> batch [-f file] [-m] >>> while executing >>> "exec batch -q long.q ${filebase}_script.sh" >>> ("probtrackx" arm line 144) >>> invoked from within >>> "switch -- $probtrack(tool) { >>> eddy_current { >>> global eddy >>> >>> set errorStr "" >>> if { $eddy(input) == "" } { set errorStr "You need to >>> specify..." >>> (procedure "fdt:apply" line 5) >>> invoked from within >>> "fdt:apply .fdt keep" >>> invoked from within >>> ".fdt.apply invoke" >>> ("uplevel" body line 1) >>> invoked from within >>> "uplevel #0 [list $w invoke]" >>> (procedure "tk::ButtonUp" line 22) >>> invoked from within >>> "tk::ButtonUp .fdt.apply" >>> (command bound to event) > ========================================================================= Date: Wed, 10 Oct 2007 16:08:00 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Sheyna Gifford <[log in to unmask]> Subject: Re: DICOM converter In-Reply-To: <[log in to unmask]> Mime-version: 1.0 Content-type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit Carlos- I highly recommend using Afni's 3dAFNItoNIFTI call. Works like a charm. Afni is free and runs on basically any platform. The only hitch: the data has to be in Afni format first. The command to convert dicoms into Afni briks is called to3d. See the docs. It's simple and easy once you've tried it once or twice. http://afni.nimh.nih.gov/afni/doc/edu/afni02_to3d http://afni.nimh.nih.gov/pub/dist/doc/program_help/to3d.html - seg On 10/10/07 2:20 PM, "Carlos Faraco" <[log in to unmask]> wrote: > I have actually tried the DICOM converter from SPM5 and it says it cannot > determine what order the images is in and will just guess. Thank you, and > everyone else, for the suggestions. I will try them. > > -Carlos -- Sheyna Gifford Research Specialist Magnetic Resonance Imaging Research Facility Brown University Box G-L 401-863-6513 ========================================================================= Date: Wed, 10 Oct 2007 21:06:00 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mahinda Yogarajah <[log in to unmask]> Subject: Robust Means MIME-Version: 1.0 Content-Type: multipart/mixed; boundary="------------040200090903040207010703" This is a multi-part message in MIME format. --------------040200090903040207010703 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Dear List/Expert Users, Could someone tell me how I can use fslstats/fslmaths to calculate robust means (eg mean between 5th and 95th percentile, or 10th and 90th centile etc) ? I know fslstats can give a robust range, but are there ways to manipulate this further ? Also is there q quicj way Thanks. Mahinda PhD student, ION --------------040200090903040207010703 Content-Type: text/x-vcard; charset=utf-8; name="m.yogarajah.vcf" Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename="m.yogarajah.vcf" begin:vcard fn:Dr M Yogarajah n:Yogarajah;Dr M org:Institute of Neurology and National Society for Epilepsy;Department of Experimental and Clinical Epilepsy email;internet:[log in to unmask] title:Clinical Research Fellow version:2.1 end:vcard --------------040200090903040207010703-- ========================================================================= Date: Wed, 10 Oct 2007 16:27:47 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Daniel Glen <[log in to unmask]> Subject: Re: DICOM converter In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Actually to3d can convert the DICOM files directly to NIFTI without using the 3dAFNItoNIFTI program by simply specifying the -prefix option to end with .nii. You can use the Dimon program with the - GERT_Reco and -dicom_org options to create a script that contains an appropriate to3d command and to generate a list of files to convert. Dimon can figure out file numbering issues caused by naming problems, duplicate or missing data. Dimon -infile_prefix s8912345/i -dicom_org -GERT_Reco -quit Daniel Glen On Oct 10, 2007, at 4:08 PM, Sheyna Gifford wrote: > Carlos- > > I highly recommend using Afni's 3dAFNItoNIFTI call. Works like a > charm. > Afni is free and runs on basically any platform. The only hitch: > the data > has to be in Afni format first. The command to convert dicoms into > Afni > briks is called to3d. > > See the docs. It's simple and easy once you've tried it once or twice. > > http://afni.nimh.nih.gov/afni/doc/edu/afni02_to3d > http://afni.nimh.nih.gov/pub/dist/doc/program_help/to3d.html > > - seg > > On 10/10/07 2:20 PM, "Carlos Faraco" <[log in to unmask]> wrote: > >> I have actually tried the DICOM converter from SPM5 and it says it >> cannot >> determine what order the images is in and will just guess. Thank >> you, and >> everyone else, for the suggestions. I will try them. >> >> -Carlos > > -- > > Sheyna Gifford > Research Specialist > Magnetic Resonance Imaging Research Facility > Brown University > Box G-L > 401-863-6513 ========================================================================= Date: Wed, 10 Oct 2007 16:45:31 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Doug Greve <[log in to unmask]> Subject: Re: DICOM converter In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit At one point the afni converter "cardinalized" the qform matrix, ie, if the matrix represented an oblique slice prescription, it would make it non-oblique. This still preserves the basic orientation (ie, left is still left and right is still right), but you end up losing scanner coordinates which can be quite useful. I know Bob is aware of the issue, but I don't know whether it is fixed or not. Anyone know? doug Daniel Glen wrote: > Actually to3d can convert the DICOM files directly to NIFTI without > using the 3dAFNItoNIFTI program by simply specifying the -prefix > option to end with .nii. You can use the Dimon program with the - > GERT_Reco and -dicom_org options to create a script that contains an > appropriate to3d command and to generate a list of files to convert. > Dimon can figure out file numbering issues caused by naming problems, > duplicate or missing data. > > Dimon -infile_prefix s8912345/i -dicom_org -GERT_Reco -quit > > Daniel Glen > > On Oct 10, 2007, at 4:08 PM, Sheyna Gifford wrote: > >> Carlos- >> >> I highly recommend using Afni's 3dAFNItoNIFTI call. Works like a >> charm. >> Afni is free and runs on basically any platform. The only hitch: the >> data >> has to be in Afni format first. The command to convert dicoms into Afni >> briks is called to3d. >> >> See the docs. It's simple and easy once you've tried it once or twice. >> >> http://afni.nimh.nih.gov/afni/doc/edu/afni02_to3d >> http://afni.nimh.nih.gov/pub/dist/doc/program_help/to3d.html >> >> - seg >> >> On 10/10/07 2:20 PM, "Carlos Faraco" <[log in to unmask]> wrote: >> >>> I have actually tried the DICOM converter from SPM5 and it says it >>> cannot >>> determine what order the images is in and will just guess. Thank >>> you, and >>> everyone else, for the suggestions. I will try them. >>> >>> -Carlos >> >> >> -- >> >> Sheyna Gifford >> Research Specialist >> Magnetic Resonance Imaging Research Facility >> Brown University >> Box G-L >> 401-863-6513 > > > -- Douglas N. Greve, Ph.D. MGH-NMR Center [log in to unmask] Phone Number: 617-724-2358 Fax: 617-726-7422 In order to help us help you, please follow the steps in: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting ========================================================================= Date: Wed, 10 Oct 2007 16:53:21 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Lokke Highstein <[log in to unmask]> Subject: Re: probtrackx GUI issue on OSX cluster In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit thanks merry, that was it, we are able to submit the probtrack jobs fine now. just a few more types of jobs to test. -lokke On Oct 10, 2007, at 2:57 PM, Merry Mani wrote: > Hi, > > I had the same problem on OS X too. I showed it to my system > administrator and he modified the fdt.tcl by changing the "exec > batch -q all.q ${filebase}_script.sh" to "exec qsub -q all.q $ > {filebase}_script.sh". Now it works fine for me. > > -Merry > > > On Oct 10, 2007, at 2:36 PM, Lokke Highstein wrote: > >> hi matthew, >> >> we don't have a long.q (or a short.q, veryshort.q, or verylong.q) >> we only are currently using the all.q >> >> so i went and changed all the references to "long.q" in the >> fdt.tcl file and edited them to all.q instead. i think that is >> what you are getting at. if you were suggesting that i remove any >> references to long.q by commenting out the whole line, i can do >> that but it seems like it would guarantee that the job wouldn't >> get submitted to the cluster then. >> >> unfortunately changing that didn't help with the GUI error, and >> even under command line use, the job still just ran on the head >> node (didn't get submitted to the SGE) >> >> here's the error output from the GUI: >> >> usage: at [-q x] [-f file] [-m] time >> at -c job [job ...] >> at [-f file] -t [[CC]YY]MMDDhhmm[.SS] >> at -r job [job ...] >> at -l -q queuename >> at -l [job ...] >> atq [-q x] [-v] >> atrm job [job ...] >> batch [-f file] [-m] >> usage: at [-q x] [-f file] [-m] time >> at -c job [job ...] >> at [-f file] -t [[CC]YY]MMDDhhmm[.SS] >> at -r job [job ...] >> at -l -q queuename >> at -l [job ...] >> atq [-q x] [-v] >> atrm job [job ...] >> batch [-f file] [-m] >> while executing >> "exec batch -q all.q ${filebase}_script.sh" >> ("probtrackx" arm line 144) >> invoked from within >> "switch -- $probtrack(tool) { >> eddy_current { >> global eddy >> >> set errorStr "" >> if { $eddy(input) == "" } { set errorStr "You need to >> specify..." >> (procedure "fdt:apply" line 5) >> invoked from within >> "fdt:apply .fdt keep" >> invoked from within >> ".fdt.apply invoke" >> ("uplevel" body line 1) >> invoked from within >> "uplevel #0 [list $w invoke]" >> (procedure "tk::ButtonUp" line 22) >> invoked from within >> "tk::ButtonUp .fdt.apply" >> (command bound to event) >> >> On Oct 10, 2007, at 6:42 AM, Matthew Webster wrote: >> >>> Hi, >>> It looks like the problem is with the batch call - the "-q" >>> option doesn't seem to be available for the osx batch command. If >>> you remove "-q long.q" from the batch command in fdt.tcl, I think >>> it should batch correctly. >>> >>> Matthew >>>> hello list, >>>> >>>> sorry to keep bugging you all, but we are making a lot of >>>> progress with our cluster, thanks to this list, and we are >>>> almost fully functional which is very exciting, as we are seeing >>>> some very impressive speeds on the larger jobs (especially >>>> bedpost!) >>>> >>>> i implemented the symbolic links which were suggested to me in >>>> earlier posts, and now we can run bedpostx jobs and FEAT jobs >>>> fine from the gui, however we seem to have hit a snag with >>>> probtrackx. here is the information i got from the user in >>>> question, please advise what the problem could be, as it works >>>> fine from the command line (although for some reason this job >>>> gets run on the head node only and never gets submitted through >>>> the qsub process to the cluster.) >>>> >>>> Running probtrackx from the GUI creates the appropriate output >>>> directory, but only contains fdt_log.tcl and fdt_script.sh >>>> >>>> If I open fdt_script.sh and paste only the probtrackx command >>>> into the terminal command line, the job runs perfectly. This is >>>> the command that I paste: >>>> >>>> /common/fsl/bin/probtrackx --mode=seedmask -x /Volumes/Clinical/ >>>> homes/tyanagihara/sweat/DTI/mask-dump/seed-internalCapsule- >>>> mask.hdr -l -c 0.2 -S 2000 --steplength=0.5 -P 5000 --stop=/ >>>> Volumes/Clinical/homes/tyanagihara/sweat/DTI/mask-dump/target1- >>>> mask.hdr --forcedir --opd -s >>>> >>>> This is the error I get when running from the GUI: >>>> usage: at [-q x] [-f file] [-m] time >>>> at -c job [job ...] >>>> at [-f file] -t [[CC]YY]MMDDhhmm[.SS] >>>> at -r job [job ...] >>>> at -l -q queuename >>>> at -l [job ...] >>>> atq [-q x] [-v] >>>> atrm job [job ...] >>>> batch [-f file] [-m] >>>> usage: at [-q x] [-f file] [-m] time >>>> at -c job [job ...] >>>> at [-f file] -t [[CC]YY]MMDDhhmm[.SS] >>>> at -r job [job ...] >>>> at -l -q queuename >>>> at -l [job ...] >>>> atq [-q x] [-v] >>>> atrm job [job ...] >>>> batch [-f file] [-m] >>>> while executing >>>> "exec batch -q long.q ${filebase}_script.sh" >>>> ("probtrackx" arm line 144) >>>> invoked from within >>>> "switch -- $probtrack(tool) { >>>> eddy_current { >>>> global eddy >>>> >>>> set errorStr "" >>>> if { $eddy(input) == "" } { set errorStr "You need to >>>> specify..." >>>> (procedure "fdt:apply" line 5) >>>> invoked from within >>>> "fdt:apply .fdt keep" >>>> invoked from within >>>> ".fdt.apply invoke" >>>> ("uplevel" body line 1) >>>> invoked from within >>>> "uplevel #0 [list $w invoke]" >>>> (procedure "tk::ButtonUp" line 22) >>>> invoked from within >>>> "tk::ButtonUp .fdt.apply" >>>> (command bound to event) >> ========================================================================= Date: Wed, 10 Oct 2007 16:56:16 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Daniel Glen <[log in to unmask]> Subject: Re: DICOM converter In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit The oblique transformation matrix is computed from the DICOM header and kept now. The matrix is stored in the AFNI header or in the sform fields of a NIFTI file. The transformation matrix can be applied to files to either "oblique" or "deoblique" a dataset using 3dWarp. Otherwise, the cardinal matrix is used in AFNI processing and display. This web page discusses how to process oblique datasets with AFNI. http://afni.nimh.nih.gov/sscc/dglen/Obliquity Daniel On Oct 10, 2007, at 4:45 PM, Doug Greve wrote: > At one point the afni converter "cardinalized" the qform matrix, > ie, if the matrix represented an oblique slice prescription, it > would make it non-oblique. This still preserves the basic > orientation (ie, left is still left and right is still right), but > you end up losing scanner coordinates which can be quite useful. I > know Bob is aware of the issue, but I don't know whether it is > fixed or not. Anyone know? > > doug > > > > > Daniel Glen wrote: > >> Actually to3d can convert the DICOM files directly to NIFTI >> without using the 3dAFNItoNIFTI program by simply specifying the - >> prefix option to end with .nii. You can use the Dimon program >> with the - GERT_Reco and -dicom_org options to create a script >> that contains an appropriate to3d command and to generate a list >> of files to convert. Dimon can figure out file numbering issues >> caused by naming problems, duplicate or missing data. >> >> Dimon -infile_prefix s8912345/i -dicom_org -GERT_Reco -quit >> >> Daniel Glen >> >> On Oct 10, 2007, at 4:08 PM, Sheyna Gifford wrote: >> >>> Carlos- >>> >>> I highly recommend using Afni's 3dAFNItoNIFTI call. Works like >>> a charm. >>> Afni is free and runs on basically any platform. The only hitch: >>> the data >>> has to be in Afni format first. The command to convert dicoms >>> into Afni >>> briks is called to3d. >>> >>> See the docs. It's simple and easy once you've tried it once or >>> twice. >>> >>> http://afni.nimh.nih.gov/afni/doc/edu/afni02_to3d >>> http://afni.nimh.nih.gov/pub/dist/doc/program_help/to3d.html >>> >>> - seg >>> >>> On 10/10/07 2:20 PM, "Carlos Faraco" <[log in to unmask]> wrote: >>> >>>> I have actually tried the DICOM converter from SPM5 and it says >>>> it cannot >>>> determine what order the images is in and will just guess. >>>> Thank you, and >>>> everyone else, for the suggestions. I will try them. >>>> >>>> -Carlos >>> >>> >>> -- >>> >>> Sheyna Gifford >>> Research Specialist >>> Magnetic Resonance Imaging Research Facility >>> Brown University >>> Box G-L >>> 401-863-6513 >> >> >> > > -- > Douglas N. Greve, Ph.D. > MGH-NMR Center > [log in to unmask] > Phone Number: 617-724-2358 Fax: 617-726-7422 > > In order to help us help you, please follow the steps in: > surfer.nmr.mgh.harvard.edu/fswiki/BugReporting ========================================================================= Date: Wed, 10 Oct 2007 22:33:30 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mark Jenkinson <[log in to unmask]> Subject: Re: Robust Means In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, You can percentiles using the -p or -P options in fslstats (depending on whether you include or exclude the zeros). This will let you calculate your specified alternatives to the robust range if you combine it with the calculator utility bc. All the best, Mark On 10 Oct 2007, at 21:06, Mahinda Yogarajah wrote: > Dear List/Expert Users, > > Could someone tell me how I can use fslstats/fslmaths to calculate > robust means (eg mean between 5th and 95th percentile, or 10th and > 90th centile etc) ? I know fslstats can give a robust range, but > are there ways to manipulate this further ? Also is there q quicj way > > Thanks. > > Mahinda > PhD student, ION > > ========================================================================= Date: Wed, 10 Oct 2007 22:40:57 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Christopher Bell <[log in to unmask]> Subject: interleaved slices and sigloss Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Hello, Has anyone tried using sigloss with interleaved slice collection? So far = I have calculated the sigloss image, and used the flirt options --inweight (sigloss output), -searchcost (normcorr) and -cost (normcorr) to register= the fmri image to our PD image. These options are successfully moving the= alignment in the right direction by about 1 mm, but there is still about = 4mm to go. I have tried registering to the T1 as well, but it seems to move t= he alignment in the opposite direction! Does anyone know if interleaved slic= e collection disrupts the sigloss calculation or have any suggestions? Than= ks. Chris Bell PS: Any suggestions on how to best apply ventricle weighting from the fmr= i to a PD or a T1, to achieve better alignment. ========================================================================= Date: Wed, 10 Oct 2007 15:20:47 -0700 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: cheng ouyang <[log in to unmask]> Subject: Blood Flow In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="0-808242306-1192054847=:86326" Content-Transfer-Encoding: 8bit --0-808242306-1192054847=:86326 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hi, FSLers, I have a question about blood flow of human brain. Normally, people like to use CBF (ml/100ml/min) to describe the blood flow. My question is that Anybody know about the approximate numbers of BLODD FlOW RATE(ml/min) in human cortex (maybe capillary, arteriole)? Cathy --------------------------------- Take the Internet to Go: Yahoo!Go puts the Internet in your pocket: mail, news, photos & more. --0-808242306-1192054847=:86326 Content-Type: text/html; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hi, FSLers,

I have a question about blood flow of human brain. Normally, people like to use CBF (ml/100ml/min) to describe the blood flow. My question is that Anybody know about the approximate numbers of BLODD FlOW RATE(ml/min) in human cortex (maybe capillary, arteriole)?


Cathy


Take the Internet to Go: Yahoo!Go puts the Internet in your pocket: mail, news, photos & more. --0-808242306-1192054847=:86326-- ========================================================================= Date: Wed, 10 Oct 2007 18:04:12 -0500 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: William Janes <[log in to unmask]> Subject: BET Extracting T1-weighted Images With Stroke Lesions Comments: cc: Binyam Nardos <[log in to unmask]> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Content-Type: text/html; charset=ISO-8859-1
All,

I am attempting multi-channel segmentation of brains = with a variety of stroke lesions. While segmentation with mfast seems painl= ess, the initial BET brain extraction is problematic.

BET does quite= well at extracting our T2-weighted images at it's default settings. The T1= -weighted images, however, do not extract cleanly. I have tried numerous it= erations, adjusting the fractional intensity threshold, using -R, -S, or -A= 2, and all result in erroneously included and excluded voxels. That is, an = unacceptable amount of grey matter is removed and a great deal of bone, mus= cle, and other tissue remains. In addition, large swaths of more shallow le= sions are excluded, defeating the purpose of the process.

The T1 and= T2 images have already been transformed to the same atlas space. Given tha= t, along with the quality of our T2 BET extraction, does the following solu= tion sound reasonable?

- Use 'bet -m' to extract the T2 and generate= a binary mask
- Use the mri=5Fmask tool from Freesurfer to apply the T2= mask to the T1 image
- Proceed with mfast using the extracted T2 and th= e masked T1

Alternatively, if this process sounds inappropriate, doe= s anyone have suggestions for better T1 extraction?

Thank you for an= y help you can provide.


William E. Janes
OTD Student
Washi= ngton University School of Medicine
Program in Occupational Therapy
C= ampus Box 8505
4444 Forest Park Avenue
St. Louis, MO 63108
(618) 9= 73-5344

= ========================================================================= Date: Wed, 10 Oct 2007 19:45:11 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matt Glasser <[log in to unmask]> Organization: ma-tea.com Subject: Re: BET Extracting T1-weighted Images With Stroke Lesions In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0058_01C80B76.127E4250" This is a multi-part message in MIME format. ------=_NextPart_000_0058_01C80B76.127E4250 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Assuming that transformation to standard space is good, you could use the brain mask of the T2 to extract the T1 (fslmaths -mas option will do this fine). However, you might also try making sure that the center of the brain extraction sphere is located in the center of the brain, and, if there is a superior/inferior bias in the extraction extent (e.g. too liberal on the top, too conservative on the bottom, you could try using the -g function. Peace, Matt. _____ From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of William Janes Sent: Wednesday, October 10, 2007 7:04 PM To: [log in to unmask] Subject: [FSL] BET Extracting T1-weighted Images With Stroke Lesions All, I am attempting multi-channel segmentation of brains with a variety of stroke lesions. While segmentation with mfast seems painless, the initial BET brain extraction is problematic. BET does quite well at extracting our T2-weighted images at it's default settings. The T1-weighted images, however, do not extract cleanly. I have tried numerous iterations, adjusting the fractional intensity threshold, using -R, -S, or -A2, and all result in erroneously included and excluded voxels. That is, an unacceptable amount of grey matter is removed and a great deal of bone, muscle, and other tissue remains. In addition, large swaths of more shallow lesions are excluded, defeating the purpose of the process. The T1 and T2 images have already been transformed to the same atlas space. Given that, along with the quality of our T2 BET extraction, does the following solution sound reasonable? - Use 'bet -m' to extract the T2 and generate a binary mask - Use the mri_mask tool from Freesurfer to apply the T2 mask to the T1 image - Proceed with mfast using the extracted T2 and the masked T1 Alternatively, if this process sounds inappropriate, does anyone have suggestions for better T1 extraction? Thank you for any help you can provide. William E. Janes OTD Student Washington University School of Medicine Program in Occupational Therapy Campus Box 8505 4444 Forest Park Avenue St. Louis, MO 63108 (618) 973-5344 ------=_NextPart_000_0058_01C80B76.127E4250 Content-Type: text/html; charset="US-ASCII" Content-Transfer-Encoding: quoted-printable

Assuming that transformation to = standard space is good, you could use the brain mask of the T2 to extract the T1 (fslmaths –mas option will do this fine).  However, you might = also try making sure that the center of the brain extraction sphere is = located in the center of the brain, and, if there is a superior/inferior bias in = the extraction extent (e.g. too liberal on the top, too conservative on the = bottom, you could try using the –g function.  =

 

Peace,

 

Matt.

 


From: = FSL - FMRIB's Software Library [mailto:[log in to unmask]] On = Behalf Of William Janes
Sent: Wednesday, October = 10, 2007 7:04 PM
To: = [log in to unmask]
Subject: [FSL] BET = Extracting T1-weighted Images With Stroke Lesions

 

All,

I am attempting multi-channel segmentation of brains with a variety of = stroke lesions. While segmentation with mfast seems painless, the initial BET = brain extraction is problematic.

BET does quite well at extracting our T2-weighted images at it's default settings. The T1-weighted images, however, do not extract cleanly. I = have tried numerous iterations, adjusting the fractional intensity threshold, using = -R, -S, or -A2, and all result in erroneously included and excluded voxels. = That is, an unacceptable amount of grey matter is removed and a great deal of = bone, muscle, and other tissue remains. In addition, large swaths of more = shallow lesions are excluded, defeating the purpose of the process.

The T1 and T2 images have already been transformed to the same atlas = space. Given that, along with the quality of our T2 BET extraction, does the = following solution sound reasonable?

- Use 'bet -m' to extract the T2 and generate a binary mask
- Use the mri_mask tool from Freesurfer to apply the T2 mask to the T1 = image
- Proceed with mfast using the extracted T2 and the masked T1

Alternatively, if this process sounds inappropriate, does anyone have suggestions for better T1 extraction?

Thank you for any help you can provide.


William E. Janes
OTD Student
Washington University School of Medicine
Program in Occupational Therapy
Campus Box = 8505
4444 Forest Park = Avenue
St. Louis, = MO 63108
(618) 973-5344

------=_NextPart_000_0058_01C80B76.127E4250-- ========================================================================= Date: Wed, 10 Oct 2007 19:32:25 -0500 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: William Janes <[log in to unmask]> Subject: Re: BET Extracting T1-weighted Images With Stroke Lesions In-Reply-To: [log in to unmask]> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Content-Type: text/html; charset=ISO-8859-1
Matt,

Thanks for the suggestions. I should have inclu= ded centering the sphere under "methods I've tried without success." I don'= t see a bias in the extraction, it seems similarly poor around the entire p= erimeter of the brain. I'll give fslmaths a shot. That sounds like a more s= imple solution than loading things into the Freesurfer framework unnecessar= ily.

Thanks again,
Bill

-----FSL -= FMRIB's Software Library [log in to unmask]"><[log in to unmask]> wrote: -----

To: [log in to unmask] C.UK
From: Matt Glasser [log in to unmask]"><[log in to unmask]>
Sent by: FSL - FMRIB'= s Software Library <[log in to unmask]>
Date: 10/10/2007 06:45PMSubject: Re: [FSL] BET Extracting T1-weighted Images With Stroke Lesions

Assuming that = transformation to standard space is good, you could use the brain mask of the T2 to extract the T1 (fslmaths –mas option will do this fine).  However, you might al= so try making sure that the center of the brain extraction sphere is located in the center of the brain, and, if there is a superior/inferior bias in the extraction extent (e.g. too liberal on the top, too conservative on the bot= tom, you could try using the –g function. =20

 

Peace,

 

Matt.

 


From: =20 FSL - FMRIB's Software Library :PersonName [mail= to:[log in to unmask]]=20 On Behalf Of=20 William Janes
Sent: Wednesday, October 10, 2007 7:04 PM
To: [log in to unmask]
Subject: [FSL] BET Extracting T1-weighted Images With Stroke Lesions

 

All,
I am attempting multi-channel segmentation of brains with a variety of stro= ke lesions. While segmentation with mfast seems painless, the initial BET brain extraction is problematic.

BET does quite well at extracting our T2-weighted images at it's default settings. The T1-weighted images, however, do not extract cleanly. I have t= ried numerous iterations, adjusting the fractional intensity threshold, using -R, -S, or -A2, and all result in erroneously included and excluded voxels. That is, an unacceptable amount of grey matter is removed and a great deal of bo= ne, muscle, and other tissue remains. In addition, large swaths of more shallow lesions are excluded, defeating the purpose of the process.

The T1 and T2 images have already been transformed to the same atlas space. Given that, along with the quality of our T2 BET extraction, does the follo= wing solution sound reasonable?

- Use 'bet -m' to extract the T2 and generate a binary mask
- Use the mri=5Fmask tool from Freesurfer to apply the T2 mask to the T1 im= age
- Proceed with mfast using the extracted T2 and the masked T1

Alternatively, if this process sounds inappropriate, does anyone have suggestions for better T1 extraction?

Thank you for any help you can provide.


William E. Janes
OTD Student
Washington =20 University =20 School of Medicine
Program in Occupational Therapy
Campus=20 Box 8505
4444 Forest Park Avenue
St. Louis ,=20 MO =20 63108
(618) 973-5344


= ========================================================================= Date: Thu, 11 Oct 2007 02:58:41 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Greg Burgess <[log in to unmask]> Subject: trouble using fslmaths as replacement for avwconv Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" I'm a bit confused about using the -kernel switch in fslmaths to replace = the functionality of avwconv. Basically, I was once able to enter avwmaths avg152_T1_brain_mask -roi 12 1 34 1 56 1 0 1 single_point ; avwconv -i single_point.nii.gz -s 8 -o sphere_8mm_radius.nii.gz I figured that combining the functionality in fslmaths would allow me to instead enter fslmaths MNI152_T1_2mm_brain_mask.nii.gz -roi 12 1 34 1 56 1 0 1 -kernel sphere 8 sphere_8mm_radius.nii.gz However, the resulting image contained only got a single point at 12 34 5= 6. To debug it, I also tried: fslmaths MNI152_T1_2mm_brain_mask.nii.gz -roi 12 1 34 1 56 1 0 1 single_p= oint fslmaths single_point.nii.gz -kernel sphere 8 sphere_8mm_radius.nii.gz But the resulting image still only contains a single point. I then tried = the simpler syntax for -kernel fslmaths MNI152_T1_2mm_brain_mask.nii.gz -roi 12 1 34 1 56 1 0 1 single_p= oint fslmaths single_point.nii.gz -kernel 3D cube.nii.gz But I still only get a single point. Lastly, I tried fslmaths MNI152_T1_2mm_brain_mask.nii.gz -kernel sphere 8 sphere_8mm_radius.nii.gz And here I got a value of 1 at every voxel in the brain. What gives? Can someone tell me what I'm missing? Thanks, Greg ========================================================================= Date: Wed, 10 Oct 2007 22:15:46 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Buyean Lee <[log in to unmask]> Subject: Re: Using SIENA in monkey MRIs In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="--------MB_8C9D9C3C49FE5E2_C3C_5CB5_FWM-M08.sysops.aol.com" ----------MB_8C9D9C3C49FE5E2_C3C_5CB5_FWM-M08.sysops.aol.com Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii" Hi Steve, Finally, I ran Siena with the monkey MRIs (it is just one subject); the siena_report.html looks much better in terms of brain extraction, coregistation, etc. This is what I did. 1. removed all tissues outside the skull and below the cerebellum. 2. modified the siena.sh in order to use separate BET option for two scans. Brain extraction is much better, but coregistration by FLIRT is not satisfactory. I can still see some movement in the overlap of the two coregistred MRIs. Is there any way for me to send these two MRIs to you so that you can take a look at them? I just don't know what else I can do to improve the BET and coregistration further. Thank you, Buyean -----Original Message----- From: Steve Smith <[log in to unmask]> To: [log in to unmask] Sent: Thu, 27 Sep 2007 1:20 am Subject: Re: [FSL] Using SIENA in monkey MRIs HI - you would also want to include a rasonable amount of the outer skull boundary that's near the brain boundary, as this gets used in SIENA.? ? To choose a FOV in FSLView and then reduce the image by this:? Click around and find the voxel coordinates (voxel not mm) that bound the region you want to keep.? Find the xmin and xmax values, and calculate the xsize = xmax - xmin? Same for y and z.? ? Then run? ? fslroi originalheadimage reducedimage xmin xsize ymin ysize zmin zsize? ? (do this for both timepoints separately)? ? and feed the reducedimages into SIENA.? ? ? Cheers.? ? On 26 Sep 2007, at 23:52, Buyean Lee wrote:? ? > Hi Steve,? >? > "Alternatively, you could work out how to reduce the field-of-view > to just include the brain of each time point before feeding the > output of that into SIENA. You would use FSLView on each image to > work out the field-of-view and then use fslroi to reduce the FOV of > the image."? >? > I created a whole brain mask for the animal brain using FSLView.? >? > But, it is not easy to understand how to use fslroi.? >? > fslroi ...? >? > I just don't know how to determine these parameters and when I am > supposed to use the mask file in fslroi.? >? > Would you kindly let me know how to reduce the field-of-view to > include the brain?? >? >? > Thank you,? >? > Buyean? > ----------------------------------------------------------------------> -----? > Check Out the new free AIM(R) Mail -- Unlimited storage and > industry-leading spam and email virus protection.? ? ---------------------------------------------------------------------------? Stephen M. Smith, Professor of Biomedical Engineering? Associate Director, Oxford University FMRIB Centre? ? FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK? +44 (0) 1865 222726 (fax 222717)? [log in to unmask] http://www.fmrib.ox.ac.uk/~steve? ---------------------------------------------------------------------------? ________________________________________________________________________ Check Out the new free AIM(R) Mail -- Unlimited storage and industry-leading spam and email virus protection. ----------MB_8C9D9C3C49FE5E2_C3C_5CB5_FWM-M08.sysops.aol.com Content-Transfer-Encoding: 7bit Content-Type: text/html; charset="us-ascii"
Hi Steve,

Finally, I ran Siena with the monkey MRIs (it is just one subject); the siena_report.html looks much better in terms of brain extraction, coregistation, etc.

This is what I did.

1. removed all tissues outside the skull and below the cerebellum.
2. modified the siena.sh in order to use separate BET option for two scans.

Brain extraction is much better, but coregistration by FLIRT is not satisfactory.
I can still see some movement in the overlap of the two coregistred MRIs.

Is there any way for me to send these two MRIs to you so that you can take a look at them?

I just don't know what else I can do to improve the BET and coregistration further.

Thank you,

Buyean



-----Original Message-----
From: Steve Smith <[log in to unmask]>
To: [log in to unmask]
Sent: Thu, 27 Sep 2007 1:20 am
Subject: Re: [FSL] Using SIENA in monkey MRIs

HI - you would also want to include a rasonable amount of the outer skull boundary that's near the brain boundary, as this gets used in SIENA. 
 
To choose a FOV in FSLView and then reduce the image by this: 
Click around and find the voxel coordinates (voxel not mm) that bound the region you want to keep. 
Find the xmin and xmax values, and calculate the xsize = xmax - xmin 
Same for y and z. 
 
Then run 
 
fslroi originalheadimage reducedimage xmin xsize ymin ysize zmin zsize 
 
(do this for both timepoints separately) 
 
and feed the reducedimages into SIENA. 
 
 
Cheers. 
 
On 26 Sep 2007, at 23:52, Buyean Lee wrote: 
 
> Hi Steve, 

> "Alternatively, you could work out how to reduce the field-of-view > to just include the brain of each time point before feeding the > output of that into SIENA. You would use FSLView on each image to > work out the field-of-view and then use fslroi to reduce the FOV of > the image." 

> I created a whole brain mask for the animal brain using FSLView. 

> But, it is not easy to understand how to use fslroi. 

> fslroi <input> <ouput> <xmin> <xsize> <ymin> ... 

> I just don't know how to determine these parameters and when I am > supposed to use the mask file in fslroi. 

> Would you kindly let me know how to reduce the field-of-view to > include the brain? 


> Thank you, 

> Buyean 
> ----------------------------------------------------------------------> ----- 
> Check Out the new free AIM(R) Mail -- Unlimited storage and > industry-leading spam and email virus protection. 
 
--------------------------------------------------------------------------- 
Stephen M. Smith, Professor of Biomedical Engineering 
Associate Director, Oxford University FMRIB Centre 
 
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK 
+44 (0) 1865 222726 (fax 222717) 
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve 
--------------------------------------------------------------------------- 

Check Out the new free AIM(R) Mail -- Unlimited storage and industry-leading spam and email virus protection.
----------MB_8C9D9C3C49FE5E2_C3C_5CB5_FWM-M08.sysops.aol.com-- ========================================================================= Date: Wed, 10 Oct 2007 22:35:28 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Buyean Lee <[log in to unmask]> Subject: Re: FLIRT: Preserver what? In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="--------MB_8C9D9C68583D7A2_C3C_5ED4_FWM-M08.sysops.aol.com" ----------MB_8C9D9C68583D7A2_C3C_5ED4_FWM-M08.sysops.aol.com Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii" Hi Steve, Do you mean FSL normalization will not preserve the concentration? I am specifically paying attention to this issue due to the following reason. I used to normalize the binding potential images of my PET images using SPM2(5) in which it preserve concentration. Since I am not satisfied with the location of the subcortical regions after normalization, I am looking for a method which can improve the spatial normalization of the subcortical regions (if you know one, let me know). ? I thought the spatial normalization used in FIRST (i.e., first_flirt) is promising. It seems to work better.? But, I don't know if 'first_flirt' will preserver concentration or not. Thank you, Buyean -----Original Message----- From: Steve Smith <[log in to unmask]> To: [log in to unmask] Sent: Mon, 8 Oct 2007 11:04 pm Subject: Re: [FSL] FLIRT: Preserver what? Hi,? ? FLIRT applies an affine-only transformation. If you wish to divide the output image by the amount of volumetric scaling (to preserve "concentration") you can extract the volumetric scaling factor using avscale and then apply this easily with the -mul option inside fslmaths.? ? Cheers, Steve.? ? On 8 Oct 2007, at 22:59, Buyean Lee wrote:? ? > Dear FSL users,? >? > When one normalizes a image in SPM2(5), one can choose which > parameter (e.g., concentration or total (=amount)) is going to be > preserved.? >? > I read the FSL website and checked FLIRT options, but I can't find > a similar option.? >? > Does FLIRT preserve the concentration?? >? > Thank you,? >? > Buyean? > Check Out the new free AIM(R) Mail -- Unlimited storage and > industry-leading spam and email virus protection.? ? ---------------------------------------------------------------------------? Stephen M. Smith, Professor of Biomedical Engineering? Associate Director, Oxford University FMRIB Centre? ? FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK? +44 (0) 1865 222726 (fax 222717)? [log in to unmask] http://www.fmrib.ox.ac.uk/~steve? ---------------------------------------------------------------------------? ________________________________________________________________________ Check Out the new free AIM(R) Mail -- Unlimited storage and industry-leading spam and email virus protection. ----------MB_8C9D9C68583D7A2_C3C_5ED4_FWM-M08.sysops.aol.com Content-Transfer-Encoding: 7bit Content-Type: text/html; charset="us-ascii"
Hi Steve,

Do you mean FSL normalization will not preserve the concentration?

I am specifically paying attention to this issue due to the following reason.

I used to normalize the binding potential images of my PET images using SPM2(5) in which it preserve concentration.
Since I am not satisfied with the location of the subcortical regions after normalization, I am looking for a method which can improve the spatial normalization of the subcortical regions (if you know one, let me know).
 
I thought the spatial normalization used in FIRST (i.e., first_flirt) is promising.
It seems to work better.  But, I don't know if 'first_flirt' will preserver concentration or not.

Thank you,

Buyean


-----Original Message-----
From: Steve Smith <[log in to unmask]>
To: [log in to unmask]
Sent: Mon, 8 Oct 2007 11:04 pm
Subject: Re: [FSL] FLIRT: Preserver what?

Hi, 
 
FLIRT applies an affine-only transformation. If you wish to divide the output image by the amount of volumetric scaling (to preserve "concentration") you can extract the volumetric scaling factor using avscale and then apply this easily with the -mul option inside fslmaths. 
 
Cheers, Steve. 
 
On 8 Oct 2007, at 22:59, Buyean Lee wrote: 
 
> Dear FSL users, 

> When one normalizes a image in SPM2(5), one can choose which > parameter (e.g., concentration or total (=amount)) is going to be > preserved. 

> I read the FSL website and checked FLIRT options, but I can't find > a similar option. 

> Does FLIRT preserve the concentration? 

> Thank you, 

> Buyean 
> Check Out the new free AIM(R) Mail -- Unlimited storage and > industry-leading spam and email virus protection. 
 
--------------------------------------------------------------------------- 
Stephen M. Smith, Professor of Biomedical Engineering 
Associate Director, Oxford University FMRIB Centre 
 
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK 
+44 (0) 1865 222726 (fax 222717) 
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve 
--------------------------------------------------------------------------- 

Check Out the new free AIM(R) Mail -- Unlimited storage and industry-leading spam and email virus protection.
----------MB_8C9D9C68583D7A2_C3C_5ED4_FWM-M08.sysops.aol.com-- ========================================================================= Date: Thu, 11 Oct 2007 08:32:33 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mark Jenkinson <[log in to unmask]> Subject: Re: trouble using fslmaths as replacement for avwconv In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Dear Greg, The kernel option only specifies what the kernel will be, not what to do with it. You also need to tell fslmaths what operation should be performed (i.e. convolution, morphological filtering, etc.) With the commands you've got at the moment it isn't actually doing any filtering - hence why you only get the single point output. For the straightforward convolution (averaging within the kernel) you want to use -fmean as the appropriate operation. If you wanted to use the spherical kernel for erosion or dilation instead, then you could use -ero or -dilF. See the collection of operations under "spatial filtering operations" for other options. All the best, Mark On 11 Oct 2007, at 02:58, Greg Burgess wrote: > I'm a bit confused about using the -kernel switch in fslmaths to > replace the > functionality of avwconv. Basically, I was once able to enter > > avwmaths avg152_T1_brain_mask -roi 12 1 34 1 56 1 0 1 single_point ; > avwconv -i single_point.nii.gz -s 8 -o sphere_8mm_radius.nii.gz > > I figured that combining the functionality in fslmaths would allow > me to > instead enter > > fslmaths MNI152_T1_2mm_brain_mask.nii.gz -roi 12 1 34 1 56 1 0 1 - > kernel > sphere 8 sphere_8mm_radius.nii.gz > > However, the resulting image contained only got a single point at > 12 34 56. > To debug it, I also tried: > > fslmaths MNI152_T1_2mm_brain_mask.nii.gz -roi 12 1 34 1 56 1 0 1 > single_point > fslmaths single_point.nii.gz -kernel sphere 8 sphere_8mm_radius.nii.gz > > But the resulting image still only contains a single point. I then > tried the > simpler syntax for -kernel > > fslmaths MNI152_T1_2mm_brain_mask.nii.gz -roi 12 1 34 1 56 1 0 1 > single_point > fslmaths single_point.nii.gz -kernel 3D cube.nii.gz > > But I still only get a single point. Lastly, I tried > fslmaths MNI152_T1_2mm_brain_mask.nii.gz -kernel sphere 8 > sphere_8mm_radius.nii.gz > > And here I got a value of 1 at every voxel in the brain. > > What gives? Can someone tell me what I'm missing? > Thanks, > Greg ========================================================================= Date: Thu, 11 Oct 2007 11:04:24 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: interleaved slices and sigloss In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - no, as long as phase encoding is always in the same direction (it will be) it doesn't matter that your slice ordering is interleaved. Have you tried doing full unwarping of the FMRI data, and not just FLIRT cost-function weighting? You should be able to easily use fslmaths to get a CSF-related weighting image. Cheers. On 10 Oct 2007, at 22:40, Christopher Bell wrote: > Hello, > > Has anyone tried using sigloss with interleaved slice collection? > So far I > have calculated the sigloss image, and used the flirt options -- > inweight > (sigloss output), -searchcost (normcorr) and -cost (normcorr) to > register > the fmri image to our PD image. These options are successfully > moving the > alignment in the right direction by about 1 mm, but there is still > about 4mm > to go. I have tried registering to the T1 as well, but it seems to > move the > alignment in the opposite direction! Does anyone know if > interleaved slice > collection disrupts the sigloss calculation or have any > suggestions? Thanks. > > Chris Bell > > PS: Any suggestions on how to best apply ventricle weighting from > the fmri > to a PD or a T1, to achieve better alignment. ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Thu, 11 Oct 2007 11:06:40 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: BET Extracting T1-weighted Images With Stroke Lesions In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, If you have the T2 and T1 aligned, and generate a brain mask from the T2 image, you can easily apply this to the T1 just using fslmaths: fslmaths T1 -mas T2-brain-mask T1_brain Cheers. On 11 Oct 2007, at 00:04, William Janes wrote: > All, > > I am attempting multi-channel segmentation of brains with a variety > of stroke lesions. While segmentation with mfast seems painless, > the initial BET brain extraction is problematic. > > BET does quite well at extracting our T2-weighted images at it's > default settings. The T1-weighted images, however, do not extract > cleanly. I have tried numerous iterations, adjusting the fractional > intensity threshold, using -R, -S, or -A2, and all result in > erroneously included and excluded voxels. That is, an unacceptable > amount of grey matter is removed and a great deal of bone, muscle, > and other tissue remains. In addition, large swaths of more shallow > lesions are excluded, defeating the purpose of the process. > > The T1 and T2 images have already been transformed to the same > atlas space. Given that, along with the quality of our T2 BET > extraction, does the following solution sound reasonable? > > - Use 'bet -m' to extract the T2 and generate a binary mask > - Use the mri_mask tool from Freesurfer to apply the T2 mask to the > T1 image > - Proceed with mfast using the extracted T2 and the masked T1 > > Alternatively, if this process sounds inappropriate, does anyone > have suggestions for better T1 extraction? > > Thank you for any help you can provide. > > > William E. Janes > OTD Student > Washington University School of Medicine > Program in Occupational Therapy > Campus Box 8505 > 4444 Forest Park Avenue > St. Louis, MO 63108 > (618) 973-5344 > ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Thu, 11 Oct 2007 11:15:32 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mark Jenkinson <[log in to unmask]> Subject: Re: FLIRT: Preserver what? In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, As I understand it, "concentration" is about local intensity so that conserving concentration when you are scaling an image (making the structure bigger or smaller) will keep the intensity values within that region the same. By comparison, conserving "total (amount)" tries to keep the product of intensity and volume the same, so that if the volume increases then the intensity within will decrease. The behaviour of FLIRT is always to keep the intensity the same, thus this would be "conserving concentration". So you should be fine to use either flirt, or scripts that call flirt, such as first_flirt. Note that for FMRI analyses, it is common to do grand-mean intensity normalisation which then rescales the intensities to have a fixed overall mean across all dimensions (space and time). If such a scaling is done then it doesn't matter if you tried to correct for the original flirt scaling or not, as it would be constant across space and time and therefore would be factored out by this normalisation. Only if you are wanting to use the intensities in something like VBM would it make a difference whether you tried to correct for the scaling (preserving "total"). Hope this helps. All the best, Mark On 11 Oct 2007, at 03:35, Buyean Lee wrote: > Hi Steve, > > Do you mean FSL normalization will not preserve the concentration? > > I am specifically paying attention to this issue due to the > following reason. > > I used to normalize the binding potential images of my PET images > using SPM2(5) in which it preserve concentration. > Since I am not satisfied with the location of the subcortical > regions after normalization, I am looking for a method which can > improve the spatial normalization of the subcortical regions (if > you know one, let me know). > > I thought the spatial normalization used in FIRST (i.e., > first_flirt) is promising. > It seems to work better. But, I don't know if 'first_flirt' will > preserver concentration or not. > > Thank you, > > Buyean > > > -----Original Message----- > From: Steve Smith <[log in to unmask]> > To: [log in to unmask] > Sent: Mon, 8 Oct 2007 11:04 pm > Subject: Re: [FSL] FLIRT: Preserver what? > > Hi, > > FLIRT applies an affine-only transformation. If you wish to divide > the output image by the amount of volumetric scaling (to preserve > "concentration") you can extract the volumetric scaling factor > using avscale and then apply this easily with the -mul option > inside fslmaths. > > Cheers, Steve. > > On 8 Oct 2007, at 22:59, Buyean Lee wrote: > > > Dear FSL users, > > > > When one normalizes a image in SPM2(5), one can choose which > > parameter (e.g., concentration or total (=amount)) is going to be > > preserved. > > > > I read the FSL website and checked FLIRT options, but I can't > find > a similar option. > > > > Does FLIRT preserve the concentration? > > > > Thank you, > > > > Buyean > > Check Out the new free AIM(R) Mail -- Unlimited storage and > > industry-leading spam and email virus protection. > > ---------------------------------------------------------------------- > ----- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > ---------------------------------------------------------------------- > ----- > Check Out the new free AIM(R) Mail -- Unlimited storage and > industry-leading spam and email virus protection. ========================================================================= Date: Thu, 11 Oct 2007 11:32:39 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: FSL on XGrid w/ XServe (now SGE) In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, Most of this has now been answered by others: On 9 Oct 2007, at 20:17, Jeremy Bronson wrote: > Hi, > > I tried to get FSL working with XGrid during the summer, but wasn't > quite successful, and was recommended to use SGE in its place by > Steve and > others. I'm giving it another go, but I've run into a few questions > while setting up SGE. Hopefully they're easy enough to answer... > > -The SGE manual says that each host must have the same user account > names and passwords, which is the alternate to XGrid's nobody:nobody > permissions, but seems equally impractical. If SGE is really > designed to be > implemented in large grids, I'm not sure how it could scale beyond a > handful of managed machines. My question is whether this is how the > lab is configured at Oxford or any other site, or alternately are all > jobs submitted under a single username? > > -All scan data resides on an XServe, automounted via NFS. What kind of > permissions are necessary on this share? > > -Is anyone using the multiple queues available in SGE, or only using > one for large jobs? We use multiple queues - but that's just a matter of choice, and either option is easy to setup. fsl_sub just needs customising to reflect whatever you setup. Cheers. > My basic goal is to allow users to login, start multiple large jobs on > the grid, then log out and retrieve the results later. The FEAT > first-level analyses tend to tie up all the machines for a time, > so I'd love > to speed them up by running the jobs in parallel and in the > background. > > > Thanks in advance! > > Jeremy > > > > > > > On Tue, 14 Aug 2007 06:24:00 +0100, Steve Smith <[log in to unmask]> > wrote: >> Hi Jeremy, >> >> I agree with Andrew, most people seem much happier with SGE than with > >> XGrid, so if it's not too late I would consider that. >> >> Anyway - yes, hopefully FSL 4.0 should be fairly easy to setup with >> either (though much easier with SGE as that's what we have so should > >> need much less customising). >> >> First, see the brief intro at: >> http://www.fmrib.ox.ac.uk/fsl/fsl/downloading.html#sge >> So if the user runs any of these programs on a machine which can >> submit to the cluster then that should happen automatically, after >> the sysadmin has: >> >> Setup the central cluster-controlling script >> $FSLDIR/bin/fsl_sub >> >> This is a heavily commented shell script and hopefully should be >> reasonably easy to follow and customise..... >> >> >> So hopefully the user can just run FSL programs on their normal >> working machine, and if it's setup to be a cluster submission host, >> then whenever a "big" job is run that will automatically get sent to > >> the queue. >> >> Cheers, Steve. >> >> >> >> >> >> On 14 Aug 2007, at 05:59, Jeremy Bronson wrote: >> >>> Hi All, >>> >>> I'm the new administrator of a neuroimaging research lab, and I'm >>> working on getting FSL and other MRI-analysis tools running on >>> XGrid. I'm not yet intimately familiar with how FSL works, so I'm > >>> hoping there are others out there who have already figured out how > >>> to run it on a cluster, so I don't have to reinvent the wheel. >>> I've heard of several existing clusters that run FSL, albeit a >>> modified version. I'm hoping it can be done with the off-the-shelf > >>> version, perhaps it's now possible with the just-released 4.0? If > >>> anybody could point me in the direction of some specifics, I'd be >>> most grateful. >>> >>> We've got an XServe with RAID that houses all the data, and FSL is > >>> installed and configured on all host (agent) machines. Most users > >>> use the GUI, and manually point the tools to the appropriate data, > >>> so I assume they'll need to familiarize themselves with the > command- >>> line tools and specifying data directories on the CLI (or via >>> GridStuffer). I'm thinking that each machine will need the data >>> volume to be auto-mounted (NFS?) at startup with appropriate >>> directories having read/write access for the 'nobody' user. Does >>> this sound correct? >>> >>> Additionally, each tool that's part of the FSL package seems to >>> launch a number of other UNIX commands during analysis, to copy, >>> move and otherwise manipulate the result data. Will this confuse >>> XGrid, or will the job and all sub-commands run until the original > >>> command completes? (e.g. A complete FEAT analysis) >>> >>> Hopefully this isn't too difficult, and afterwards I'd like to take > >>> the time to post a HOWTO on macresearch.org or the like, so others > >>> might take advantage of the info. Thanks in advance to anyone who > >>> might be able to help. >>> >>> >>> Jeremy Bronson >>> Systems Administrator >>> Frey Research Lab >>> University of Oregon >> >> >> > > ---------------------------------------------------------------------- > -- >> --- >> Stephen M. Smith, Professor of Biomedical Engineering >> Associate Director, Oxford University FMRIB Centre >> >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >> +44 (0) 1865 222726 (fax 222717) >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >> > > ---------------------------------------------------------------------- > -- >> --- >> > > > > > > Hi, > > I tried to get FSL working with XGrid during the summer, but wasn't > quite successful, and was recommended to use SGE in its place by > Steve and > others. I'm giving it another go, but I've run into a few questions > while setting up SGE. Hopefully they're easy enough to answer... > > -The SGE manual says that each host must have the same user account > names and passwords, which is the alternate to XGrid's nobody:nobody > permissions, but seems equally impractical. If SGE is really > designed to be > implemented in large grids, I'm not sure how it could scale beyond a > handful of managed machines. My question is whether this is how the > lab is configured at Oxford or any other site, or alternately are all > jobs submitted under a single username? > > -All scan data resides on an XServe, automounted via NFS. What kind of > permissions are necessary on this share? > > -Is anyone using the multiple queues available in SGE, or only using > one for large jobs? > > My basic goal is to allow users to login, start multiple large jobs on > the grid, then log out and retrieve the results later. The FEAT > first-level analyses tend to tie up all the machines for a time, > so I'd love > to speed them up by running the jobs in parallel and in the > background. > > > Thanks in advance! > > Jeremy > > > > > > > On Tue, 14 Aug 2007 06:24:00 +0100, Steve Smith <[log in to unmask]> > wrote: >> Hi Jeremy, >> >> I agree with Andrew, most people seem much happier with SGE than with > >> XGrid, so if it's not too late I would consider that. >> >> Anyway - yes, hopefully FSL 4.0 should be fairly easy to setup with >> either (though much easier with SGE as that's what we have so should > >> need much less customising). >> >> First, see the brief intro at: >> http://www.fmrib.ox.ac.uk/fsl/fsl/downloading.html#sge >> So if the user runs any of these programs on a machine which can >> submit to the cluster then that should happen automatically, after >> the sysadmin has: >> >> Setup the central cluster-controlling script >> $FSLDIR/bin/fsl_sub >> >> This is a heavily commented shell script and hopefully should be >> reasonably easy to follow and customise..... >> >> >> So hopefully the user can just run FSL programs on their normal >> working machine, and if it's setup to be a cluster submission host, >> then whenever a "big" job is run that will automatically get sent to > >> the queue. >> >> Cheers, Steve. >> >> >> >> >> >> On 14 Aug 2007, at 05:59, Jeremy Bronson wrote: >> >>> Hi All, >>> >>> I'm the new administrator of a neuroimaging research lab, and I'm >>> working on getting FSL and other MRI-analysis tools running on >>> XGrid. I'm not yet intimately familiar with how FSL works, so I'm > >>> hoping there are others out there who have already figured out how > >>> to run it on a cluster, so I don't have to reinvent the wheel. >>> I've heard of several existing clusters that run FSL, albeit a >>> modified version. I'm hoping it can be done with the off-the-shelf > >>> version, perhaps it's now possible with the just-released 4.0? If > >>> anybody could point me in the direction of some specifics, I'd be >>> most grateful. >>> >>> We've got an XServe with RAID that houses all the data, and FSL is > >>> installed and configured on all host (agent) machines. Most users > >>> use the GUI, and manually point the tools to the appropriate data, > >>> so I assume they'll need to familiarize themselves with the > command- >>> line tools and specifying data directories on the CLI (or via >>> GridStuffer). I'm thinking that each machine will need the data >>> volume to be auto-mounted (NFS?) at startup with appropriate >>> directories having read/write access for the 'nobody' user. Does >>> this sound correct? >>> >>> Additionally, each tool that's part of the FSL package seems to >>> launch a number of other UNIX commands during analysis, to copy, >>> move and otherwise manipulate the result data. Will this confuse >>> XGrid, or will the job and all sub-commands run until the original > >>> command completes? (e.g. A complete FEAT analysis) >>> >>> Hopefully this isn't too difficult, and afterwards I'd like to take > >>> the time to post a HOWTO on macresearch.org or the like, so others > >>> might take advantage of the info. Thanks in advance to anyone who > >>> might be able to help. >>> >>> >>> Jeremy Bronson >>> Systems Administrator >>> Frey Research Lab >>> University of Oregon >> >> >> > > ---------------------------------------------------------------------- > -- >> --- >> Stephen M. Smith, Professor of Biomedical Engineering >> Associate Director, Oxford University FMRIB Centre >> >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >> +44 (0) 1865 222726 (fax 222717) >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >> > > ---------------------------------------------------------------------- > -- >> --- >> > > > > > > ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Thu, 11 Oct 2007 11:35:05 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: probtrackx GUI issue on OSX cluster In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Yes - in fact this was a bug - the call to "batch" in fdt.tcl should have been to fsl_sub and using the -T option not the explicit queue name - we'll make a patch shortly (though the direct fix mentioned to call qsub is fine too). Cheers. On 10 Oct 2007, at 21:53, Lokke Highstein wrote: > thanks merry, > > that was it, we are able to submit the probtrack jobs fine now. > > just a few more types of jobs to test. > > -lokke > On Oct 10, 2007, at 2:57 PM, Merry Mani wrote: > >> Hi, >> >> I had the same problem on OS X too. I showed it to my system >> administrator and he modified the fdt.tcl by changing the "exec >> batch -q all.q ${filebase}_script.sh" to "exec qsub -q all.q $ >> {filebase}_script.sh". Now it works fine for me. >> >> -Merry >> >> >> On Oct 10, 2007, at 2:36 PM, Lokke Highstein wrote: >> >>> hi matthew, >>> >>> we don't have a long.q (or a short.q, veryshort.q, or verylong.q) >>> we only are currently using the all.q >>> >>> so i went and changed all the references to "long.q" in the >>> fdt.tcl file and edited them to all.q instead. i think that is >>> what you are getting at. if you were suggesting that i remove >>> any references to long.q by commenting out the whole line, i can >>> do that but it seems like it would guarantee that the job >>> wouldn't get submitted to the cluster then. >>> >>> unfortunately changing that didn't help with the GUI error, and >>> even under command line use, the job still just ran on the head >>> node (didn't get submitted to the SGE) >>> >>> here's the error output from the GUI: >>> >>> usage: at [-q x] [-f file] [-m] time >>> at -c job [job ...] >>> at [-f file] -t [[CC]YY]MMDDhhmm[.SS] >>> at -r job [job ...] >>> at -l -q queuename >>> at -l [job ...] >>> atq [-q x] [-v] >>> atrm job [job ...] >>> batch [-f file] [-m] >>> usage: at [-q x] [-f file] [-m] time >>> at -c job [job ...] >>> at [-f file] -t [[CC]YY]MMDDhhmm[.SS] >>> at -r job [job ...] >>> at -l -q queuename >>> at -l [job ...] >>> atq [-q x] [-v] >>> atrm job [job ...] >>> batch [-f file] [-m] >>> while executing >>> "exec batch -q all.q ${filebase}_script.sh" >>> ("probtrackx" arm line 144) >>> invoked from within >>> "switch -- $probtrack(tool) { >>> eddy_current { >>> global eddy >>> >>> set errorStr "" >>> if { $eddy(input) == "" } { set errorStr "You need to >>> specify..." >>> (procedure "fdt:apply" line 5) >>> invoked from within >>> "fdt:apply .fdt keep" >>> invoked from within >>> ".fdt.apply invoke" >>> ("uplevel" body line 1) >>> invoked from within >>> "uplevel #0 [list $w invoke]" >>> (procedure "tk::ButtonUp" line 22) >>> invoked from within >>> "tk::ButtonUp .fdt.apply" >>> (command bound to event) >>> >>> On Oct 10, 2007, at 6:42 AM, Matthew Webster wrote: >>> >>>> Hi, >>>> It looks like the problem is with the batch call - the "-q" >>>> option doesn't seem to be available for the osx batch command. >>>> If you remove "-q long.q" from the batch command in fdt.tcl, I >>>> think it should batch correctly. >>>> >>>> Matthew >>>>> hello list, >>>>> >>>>> sorry to keep bugging you all, but we are making a lot of >>>>> progress with our cluster, thanks to this list, and we are >>>>> almost fully functional which is very exciting, as we are >>>>> seeing some very impressive speeds on the larger jobs >>>>> (especially bedpost!) >>>>> >>>>> i implemented the symbolic links which were suggested to me in >>>>> earlier posts, and now we can run bedpostx jobs and FEAT jobs >>>>> fine from the gui, however we seem to have hit a snag with >>>>> probtrackx. here is the information i got from the user in >>>>> question, please advise what the problem could be, as it works >>>>> fine from the command line (although for some reason this job >>>>> gets run on the head node only and never gets submitted through >>>>> the qsub process to the cluster.) >>>>> >>>>> Running probtrackx from the GUI creates the appropriate output >>>>> directory, but only contains fdt_log.tcl and fdt_script.sh >>>>> >>>>> If I open fdt_script.sh and paste only the probtrackx command >>>>> into the terminal command line, the job runs perfectly. This is >>>>> the command that I paste: >>>>> >>>>> /common/fsl/bin/probtrackx --mode=seedmask -x /Volumes/Clinical/ >>>>> homes/tyanagihara/sweat/DTI/mask-dump/seed-internalCapsule- >>>>> mask.hdr -l -c 0.2 -S 2000 --steplength=0.5 -P 5000 --stop=/ >>>>> Volumes/Clinical/homes/tyanagihara/sweat/DTI/mask-dump/target1- >>>>> mask.hdr --forcedir --opd -s >>>>> >>>>> This is the error I get when running from the GUI: >>>>> usage: at [-q x] [-f file] [-m] time >>>>> at -c job [job ...] >>>>> at [-f file] -t [[CC]YY]MMDDhhmm[.SS] >>>>> at -r job [job ...] >>>>> at -l -q queuename >>>>> at -l [job ...] >>>>> atq [-q x] [-v] >>>>> atrm job [job ...] >>>>> batch [-f file] [-m] >>>>> usage: at [-q x] [-f file] [-m] time >>>>> at -c job [job ...] >>>>> at [-f file] -t [[CC]YY]MMDDhhmm[.SS] >>>>> at -r job [job ...] >>>>> at -l -q queuename >>>>> at -l [job ...] >>>>> atq [-q x] [-v] >>>>> atrm job [job ...] >>>>> batch [-f file] [-m] >>>>> while executing >>>>> "exec batch -q long.q ${filebase}_script.sh" >>>>> ("probtrackx" arm line 144) >>>>> invoked from within >>>>> "switch -- $probtrack(tool) { >>>>> eddy_current { >>>>> global eddy >>>>> >>>>> set errorStr "" >>>>> if { $eddy(input) == "" } { set errorStr "You need to >>>>> specify..." >>>>> (procedure "fdt:apply" line 5) >>>>> invoked from within >>>>> "fdt:apply .fdt keep" >>>>> invoked from within >>>>> ".fdt.apply invoke" >>>>> ("uplevel" body line 1) >>>>> invoked from within >>>>> "uplevel #0 [list $w invoke]" >>>>> (procedure "tk::ButtonUp" line 22) >>>>> invoked from within >>>>> "tk::ButtonUp .fdt.apply" >>>>> (command bound to event) >>> ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Thu, 11 Oct 2007 12:41:20 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mahinda Yogarajah <[log in to unmask]> Subject: probtrackx error MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Dear FSL Experts, We are currently trying to start using probtrackx. All the appropriate files appear to be present having run bedpostx. Using either a single voxel, or a single mask and the GUI we encounter the following error message. Probtrackx appears to write out the appropriate directory name within which there are two files fdt_log.tcl and fdt_script.sh. Usage: at [-V] [-q x] [-f file] [-m] time at [-V] [-q x] [-f file] [-m] -t [[CC]YY]MMDDhhmm at -c job [job...] (atq | at -l) [-V] [-q x] (atrm | at -d | at -r) [-V] [-q x] job ... batch [-V] [-f file] [-m] Usage: at [-V] [-q x] [-f file] [-m] time at [-V] [-q x] [-f file] [-m] -t [[CC]YY]MMDDhhmm at -c job [job...] (atq | at -l) [-V] [-q x] (atrm | at -d | at -r) [-V] [-q x] job ... batch [-V] [-f file] [-m] while executing "exec batch -q long.q ${filebase}_script.sh" ("probtrackx" arm line 134) invoked from within "switch -- $probtrack(tool) { eddy_current { global eddy set errorStr "" if { $eddy(input) == "" } { set errorStr "You need to specify..." (procedure "fdt:apply" line 5) invoked from within "fdt:apply .fdt keep" invoked from within ".fdt.apply invoke" ("uplevel" body line 1) invoked from within "uplevel #0 [list $w invoke]" (procedure "tk::ButtonUp" line 22) invoked from within "tk::ButtonUp .fdt.apply" (command bound to event) Running probtrackx from the command line with only the default options present gives the following error: Log directory is: ++ 7 sl 0 sl 1 sl 2 sl 3 sl 4 and so on... to sl 38 followed by run 53 53 38 Segmentation fault Could anyone help point us in the right direction as to what may be the problem ? Thanks. Mahinda ========================================================================= Date: Thu, 11 Oct 2007 13:36:00 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "M. Pham" <[log in to unmask]> Subject: AW: [FSL] Blood Flow In-Reply-To: A<[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01C80BFA.E607D1E8" This is a multi-part message in MIME format. ------_=_NextPart_001_01C80BFA.E607D1E8 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Hi,=20 =20 an often quoted normal human CBF value is: =20 ~ 50ml / 100ml /min (to my knowledge this value is not very different in animals relevant to experimental cerebral ischemia) =20 CBF landmarks in ischemia mostly derived from animal models are: =20 ~ 10-20ml =3D=3D> ischemia associated with increasingly abnormal = neuronal function (often determined by EEG) and biochemical abnormalities (acidosis/ATP depletion) - at this level of ischemia alterations are reversible if blood flow is restored timely (concept of ischemic penumbra/corona) =20 < 10 ml =3D=3D> rapidly ensuing infarction/cell death regardless of the duration of ischemia =20 Hope this helps! Mirko ________________________________ Von: FSL - FMRIB's Software Library [mailto:[log in to unmask]] Im Auftrag von cheng ouyang Gesendet: Donnerstag, 11. Oktober 2007 00:21 An: [log in to unmask] Betreff: [FSL] Blood Flow Hi, FSLers, I have a question about blood flow of human brain. Normally, people like to use CBF (ml/100ml/min) to describe the blood flow. My question is that Anybody know about the approximate numbers of BLODD FlOW RATE(ml/min) in human cortex (maybe capillary, arteriole)? Cathy ________________________________ Take the Internet to Go: Yahoo!Go puts the Internet in your pocket: mail, news, photos & more.=20 ------_=_NextPart_001_01C80BFA.E607D1E8 Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable
Hi,
 
an often quoted normal human CBF value=20 is:
 
~ 50ml / 100ml=20 /min (to my knowledge this value is = not very=20 different in animals relevant to experimental cerebral=20 ischemia)
 
CBF landmarks in=20 ischemia mostly derived from animal models = are:
 
~ 10-20ml =3D=3D> ischemia associated with = increasingly=20 abnormal neuronal function (often determined by EEG) and biochemical=20 abnormalities (acidosis/ATP depletion) - at this level of ischemia = alterations=20 are reversible if blood flow is restored timely (concept of = ischemic=20 penumbra/corona)
 
< 10 ml =3D=3D> rapidly = ensuing infarction/cell=20 death regardless of the duration of ischemia
 
Hope this helps!
Mirko


Von: FSL - FMRIB's Software Library=20 [mailto:[log in to unmask]] Im Auftrag von cheng=20 ouyang
Gesendet: Donnerstag, 11. Oktober 2007 = 00:21
An:=20 [log in to unmask]
Betreff: [FSL] Blood = Flow

Hi, FSLers,

I have a question about blood flow of = human brain.=20 Normally, people like to use CBF (ml/100ml/min) to describe the blood = flow. My=20 question is that Anybody know about the approximate numbers of BLODD = FlOW=20 RATE(ml/min) in human cortex (maybe capillary,=20 arteriole)?


Cathy


Take the Internet to Go: Yahoo!Go puts the Internet=20 in your pocket: mail, news, photos & more. ------_=_NextPart_001_01C80BFA.E607D1E8-- ========================================================================= Date: Thu, 11 Oct 2007 12:39:00 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Saad Jbabdi <[log in to unmask]> Subject: Re: probtrackx error In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: multipart/alternative; boundary=Apple-Mail-12--952824673 --Apple-Mail-12--952824673 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed It looks like probtrackx could not read your bedpostx output. Did you specify the bedpostx directory correctly? are the files inside this directory called merged_* ? are they corrupted? Cheers, Saad. On 11 Oct 2007, at 12:41, Mahinda Yogarajah wrote: > Dear FSL Experts, > > We are currently trying to start using probtrackx. All the > appropriate files appear to be present having run bedpostx. Using > either a single voxel, or a single mask and the GUI we encounter > the following error message. Probtrackx appears to write out the > appropriate directory name within which there are two files > fdt_log.tcl and fdt_script.sh. > > Usage: at [-V] [-q x] [-f file] [-m] time > at [-V] [-q x] [-f file] [-m] -t [[CC]YY]MMDDhhmm > at -c job [job...] > (atq | at -l) [-V] [-q x] > (atrm | at -d | at -r) [-V] [-q x] job ... > batch [-V] [-f file] [-m] > Usage: at [-V] [-q x] [-f file] [-m] time > at [-V] [-q x] [-f file] [-m] -t [[CC]YY]MMDDhhmm > at -c job [job...] > (atq | at -l) [-V] [-q x] > (atrm | at -d | at -r) [-V] [-q x] job ... > batch [-V] [-f file] [-m] > while executing > "exec batch -q long.q ${filebase}_script.sh" > ("probtrackx" arm line 134) > invoked from within > "switch -- $probtrack(tool) { > eddy_current { > global eddy > > set errorStr "" > if { $eddy(input) == "" } { set errorStr "You need to > specify..." > (procedure "fdt:apply" line 5) > invoked from within > "fdt:apply .fdt keep" > invoked from within > ".fdt.apply invoke" > ("uplevel" body line 1) > invoked from within > "uplevel #0 [list $w invoke]" > (procedure "tk::ButtonUp" line 22) > invoked from within > "tk::ButtonUp .fdt.apply" > (command bound to event) > > > Running probtrackx from the command line with only the default > options present gives the following error: > Log directory is: ++ > 7 > sl 0 > sl 1 > sl 2 > sl 3 > sl 4 > and so on... to sl 38 followed by > run > 53 53 38 > Segmentation fault > > Could anyone help point us in the right direction as to what may be > the problem ? > > Thanks. > > Mahinda ------------------------------------------------------------------------ --- Saad Jbabdi, Postdoctoral Research Assistant, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222545 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~saad ------------------------------------------------------------------------ --- --Apple-Mail-12--952824673 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1
It looks like probtrackx = could not read your bedpostx output. Did you specify the bedpostx = directory correctly? are the files inside this directory called merged_* = ? are they corrupted?

Cheers,
Saad.
=


On 11 Oct = 2007, at 12:41, Mahinda Yogarajah wrote:

Dear FSL Experts,

We are = currently trying to start using probtrackx.=A0 All the appropriate files = appear to be present having run bedpostx.=A0 Using either a single voxel, = or a single mask and the GUI we encounter the following error = message.=A0 Probtrackx = appears to write out the appropriate directory name within which there = are two files fdt_log.tcl and fdt_script.sh.

Usage: = at [-V] [-q x] [-f file] [-m] time
=A0 =A0 =A0 at [-V] [-q x] [-f = file] [-m] -t [[CC]YY]MMDDhhmm
=A0 =A0 =A0 at -c job = [job...]
=A0 =A0 =A0 (atq | at -l) [-V] = [-q x]
=A0 =A0 =A0 (atrm | at -d | at = -r) [-V] [-q x] job ...
=A0 =A0 =A0 batch [-V] [-f file] = [-m]
Usage: at [-V] [-q x] [-f file] = [-m] time
=A0 =A0 =A0 at [-V] [-q x] [-f = file] [-m] -t [[CC]YY]MMDDhhmm
=A0 =A0 =A0 at -c job = [job...]
=A0 =A0 =A0 (atq | at -l) [-V] = [-q x]
=A0 =A0 =A0 (atrm | at -d | at = -r) [-V] [-q x] job ...
=A0 =A0 =A0 batch [-V] [-f file] = [-m]
=A0=A0 while executing
"exec batch -q long.q = ${filebase}_script.sh"
=A0=A0 ("probtrackx" arm line = 134)
=A0=A0 invoked from = within
"switch -- $probtrack(tool) = {
=A0=A0 = eddy_current {
=A0=A0 =A0 =A0 global = eddy

=A0=A0 =A0 =A0 = set errorStr ""
=A0=A0 =A0 =A0 if { $eddy(input) = =3D=3D "" } { set errorStr "You need to specify..."
=A0=A0 = (procedure "fdt:apply" line 5)
=A0=A0 invoked from = within
"fdt:apply .fdt keep"
=A0=A0 = invoked from within
".fdt.apply = invoke"
=A0=A0 ("uplevel" body line = 1)
=A0=A0 = invoked from within
"uplevel #0 = [list $w invoke]"
=A0=A0 (procedure "tk::ButtonUp" = line 22)
=A0=A0 invoked from = within
"tk::ButtonUp = .fdt.apply"
=A0=A0 (command bound to = event)


Running = probtrackx from the command line with only the default options present = gives the following error:
Log directory = is: ++
7
sl = 0
sl 1
sl = 2
sl 3
sl = 4
and so on... to sl 38 followed by
run
53 53 = 38
Segmentation fault

Could anyone = help point us in the right direction as to what may be the problem = ?

Thanks.

Mahinda
=

Saad Jbabdi,=A0
Postdoctoral Research Assistant,=A0=A0
Oxford University FMRIB Centre

FMRIB, = JR Hospital, Headington, Oxford=A0=A0OX3 9DU, UK
+44 (0) 1865 222545=A0=A0(fax 222717)


=

= --Apple-Mail-12--952824673-- ========================================================================= Date: Thu, 11 Oct 2007 12:45:30 +0100 Reply-To: FSL - FMRIB's Software Library <
[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Shireen Kwiatkowska-Naqvi <[log in to unmask]> Subject: Problems with siena Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain Dear Mr Smith, I'm a newbie user of fsl, and I'm encountering some problems with analysi= ng data with siena: I've=20 tried out all the possible options found in the archives and manual, but = did not come to any good=20 results (the brain extraction does not work well) I'm using Siena to interpret pre- and post training T1 data, using the -S= command cannot get rid of=20 the eye region, combining different parameters does not help either Would it be possible to send you some sample data, so you could take a lo= ok at it? With regards, Shireen Kwiatkowska-Naqvi MA Student Max-Planck-Institute for Development Berlin ========================================================================= Date: Thu, 11 Oct 2007 13:47:34 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matthew Webster <[log in to unmask]> Subject: Re: probtrackx error In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: multipart/alternative; boundary=Apple-Mail-1--948710396 --Apple-Mail-1--948710396 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Hi, This is a bug in the FDT gui, for more information see the thread: probtrackx GUI issue on OSX cluster. We'll be issuing a fix in the next patch ( potentially next week ), but for now one possible fix is to edit $FSLDIR/tcl/fdt.tcl and replace exec batch -q long.q ${filebase}_script.sh with exec fsl_sub -T 1439 ${filebase}_script.sh Regards Matthew > Dear FSL Experts, > > We are currently trying to start using probtrackx. All the > appropriate files appear to be present having run bedpostx. Using > either a single voxel, or a single mask and the GUI we encounter > the following error message. Probtrackx appears to write out the > appropriate directory name within which there are two files > fdt_log.tcl and fdt_script.sh. > > Usage: at [-V] [-q x] [-f file] [-m] time > at [-V] [-q x] [-f file] [-m] -t [[CC]YY]MMDDhhmm > at -c job [job...] > (atq | at -l) [-V] [-q x] > (atrm | at -d | at -r) [-V] [-q x] job ... > batch [-V] [-f file] [-m] > Usage: at [-V] [-q x] [-f file] [-m] time > at [-V] [-q x] [-f file] [-m] -t [[CC]YY]MMDDhhmm > at -c job [job...] > (atq | at -l) [-V] [-q x] > (atrm | at -d | at -r) [-V] [-q x] job ... > batch [-V] [-f file] [-m] > while executing > "exec batch -q long.q ${filebase}_script.sh" > ("probtrackx" arm line 134) > invoked from within > "switch -- $probtrack(tool) { > eddy_current { > global eddy > > set errorStr "" > if { $eddy(input) == "" } { set errorStr "You need to > specify..." > (procedure "fdt:apply" line 5) > invoked from within > "fdt:apply .fdt keep" > invoked from within > ".fdt.apply invoke" > ("uplevel" body line 1) > invoked from within > "uplevel #0 [list $w invoke]" > (procedure "tk::ButtonUp" line 22) > invoked from within > "tk::ButtonUp .fdt.apply" > (command bound to event) > > > Running probtrackx from the command line with only the default > options present gives the following error: > Log directory is: ++ > 7 > sl 0 > sl 1 > sl 2 > sl 3 > sl 4 > and so on... to sl 38 followed by > run > 53 53 38 > Segmentation fault > > Could anyone help point us in the right direction as to what may be > the problem ? > > Thanks. > > Mahinda --Apple-Mail-1--948710396 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1
Hi,
=A0 =A0 =A0This is a bug in the FDT gui, = for more information see the thread:=A0probtrackx GUI issue on OSX = cluster. We'll be issuing a fix in the next patch ( potentially next = week ), but for now one possible fix is to edit $FSLDIR/tcl/fdt.tcl and = replace

=A0exec batch -q long.q = ${filebase}_script.sh

with

exec fsl_sub -T 1439 = ${filebase}_script.sh

Regards

Matthew



Dear FSL Experts,

We are = currently trying to start using probtrackx.=A0 All the appropriate files = appear to be present having run bedpostx.=A0 Using either a single voxel, = or a single mask and the GUI we encounter the following error = message.=A0 Probtrackx = appears to write out the appropriate directory name within which there = are two files fdt_log.tcl and fdt_script.sh.

Usage: = at [-V] [-q x] [-f file] [-m] time
=A0 =A0 =A0 at [-V] [-q x] [-f = file] [-m] -t [[CC]YY]MMDDhhmm
=A0 =A0 =A0 at -c job = [job...]
=A0 =A0 =A0 (atq | at -l) [-V] = [-q x]
=A0 =A0 =A0 (atrm | at -d | at = -r) [-V] [-q x] job ...
=A0 =A0 =A0 batch [-V] [-f file] = [-m]
Usage: at [-V] [-q x] [-f file] = [-m] time
=A0 =A0 =A0 at [-V] [-q x] [-f = file] [-m] -t [[CC]YY]MMDDhhmm
=A0 =A0 =A0 at -c job = [job...]
=A0 =A0 =A0 (atq | at -l) [-V] = [-q x]
=A0 =A0 =A0 (atrm | at -d | at = -r) [-V] [-q x] job ...
=A0 =A0 =A0 batch [-V] [-f file] = [-m]
=A0=A0 while executing
"exec batch -q long.q = ${filebase}_script.sh"
=A0=A0 ("probtrackx" arm line = 134)
=A0=A0 invoked from = within
"switch -- $probtrack(tool) = {
=A0=A0 = eddy_current {
=A0=A0 =A0 =A0 global = eddy

=A0=A0 =A0 =A0 = set errorStr ""
=A0=A0 =A0 =A0 if { $eddy(input) = =3D=3D "" } { set errorStr "You need to specify..."
=A0=A0 = (procedure "fdt:apply" line 5)
=A0=A0 invoked from = within
"fdt:apply .fdt keep"
=A0=A0 = invoked from within
".fdt.apply = invoke"
=A0=A0 ("uplevel" body line = 1)
=A0=A0 = invoked from within
"uplevel #0 = [list $w invoke]"
=A0=A0 (procedure "tk::ButtonUp" = line 22)
=A0=A0 invoked from = within
"tk::ButtonUp = .fdt.apply"
=A0=A0 (command bound to = event)


Running = probtrackx from the command line with only the default options present = gives the following error:
Log directory = is: ++
7
sl = 0
sl 1
sl = 2
sl 3
sl = 4
and so on... to sl 38 followed by
run
53 53 = 38
Segmentation fault

Could anyone = help point us in the right direction as to what may be the problem = ?

Thanks.

Mahinda
=

= --Apple-Mail-1--948710396-- ========================================================================= Date: Thu, 11 Oct 2007 08:44:41 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Ted Yanagihara <[log in to unmask]> Subject: Re: probtrackx error In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_16077_12249417.1192106681895" ------=_Part_16077_12249417.1192106681895 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi Mahinda, In addition to checking into what Saad has suggested, did you also see the exchange between Merry, Lokke, and Steve on the list? You may be having the same problem we were with the "batch -q long.q" command. One way to make sure that you have done everything correctly to set probtrackx up to run is to run from the command line instead of the GUI. To get the correct command, open the fdt_script.sh file that was created by the failed run that you mentioned. Then copy the entire probtrackx command into the command line (this should start on the third line of the fdt_script.sh file and might roll over into the next line as well). Hope this helps! ted On 10/11/07, Saad Jbabdi <[log in to unmask]> wrote: > > It looks like probtrackx could not read your bedpostx output. Did you > specify the bedpostx directory correctly? are the files inside this > directory called merged_* ? are they corrupted? > > Cheers, > Saad. > > > On 11 Oct 2007, at 12:41, Mahinda Yogarajah wrote: > > Dear FSL Experts, > > We are currently trying to start using probtrackx. All the appropriate > files appear to be present having run bedpostx. Using either a single > voxel, or a single mask and the GUI we encounter the following error > message. Probtrackx appears to write out the appropriate directory name > within which there are two files fdt_log.tcl and fdt_script.sh. > > Usage: at [-V] [-q x] [-f file] [-m] time > at [-V] [-q x] [-f file] [-m] -t [[CC]YY]MMDDhhmm > at -c job [job...] > (atq | at -l) [-V] [-q x] > (atrm | at -d | at -r) [-V] [-q x] job ... > batch [-V] [-f file] [-m] > Usage: at [-V] [-q x] [-f file] [-m] time > at [-V] [-q x] [-f file] [-m] -t [[CC]YY]MMDDhhmm > at -c job [job...] > (atq | at -l) [-V] [-q x] > (atrm | at -d | at -r) [-V] [-q x] job ... > batch [-V] [-f file] [-m] > while executing > "exec batch -q long.q ${filebase}_script.sh" > ("probtrackx" arm line 134) > invoked from within > "switch -- $probtrack(tool) { > eddy_current { > global eddy > > set errorStr "" > if { $eddy(input) == "" } { set errorStr "You need to specify..." > (procedure "fdt:apply" line 5) > invoked from within > "fdt:apply .fdt keep" > invoked from within > ".fdt.apply invoke" > ("uplevel" body line 1) > invoked from within > "uplevel #0 [list $w invoke]" > (procedure "tk::ButtonUp" line 22) > invoked from within > "tk::ButtonUp .fdt.apply" > (command bound to event) > > > Running probtrackx from the command line with only the default options > present gives the following error: > Log directory is: ++ > 7 > sl 0 > sl 1 > sl 2 > sl 3 > sl 4 > and so on... to sl 38 followed by > run > 53 53 38 > Segmentation fault > > Could anyone help point us in the right direction as to what may be the > problem ? > > Thanks. > > Mahinda > > > > --------------------------------------------------------------------------- > Saad Jbabdi, > Postdoctoral Research Assistant, > Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222545 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~saad > > --------------------------------------------------------------------------- > > > > > ------=_Part_16077_12249417.1192106681895 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi Mahinda,

In addition to checking into what Saad has suggested, did you also see the exchange between Merry, Lokke, and Steve on the list? You may be having the same problem we were with the "batch -q long.q" command. One way to make sure that you have done everything correctly to set probtrackx up to run is to run from the command line instead of the GUI. To get the correct command, open the fdt_script.sh file that was created by the failed run that you mentioned. Then copy the entire probtrackx command into the command line (this should start on the third line of the fdt_script.sh file and might roll over into the next line as well).

Hope this helps!

ted

On 10/11/07, Saad Jbabdi <[log in to unmask]> wrote:
It looks like probtrackx could not read your bedpostx output. Did you specify the bedpostx directory correctly? are the files inside this directory called merged_* ? are they corrupted?

Cheers,
Saad.


On 11 Oct 2007, at 12:41, Mahinda Yogarajah wrote:

Dear FSL Experts,

We are currently trying to start using probtrackx.  All the appropriate files appear to be present having run bedpostx.   Using either a single voxel, or a single mask and the GUI we encounter the following error message.  Probtrackx appears to write out the appropriate directory name within which there are two files fdt_log.tcl and fdt_script.sh.

Usage: at [-V] [-q x] [-f file] [-m] time
      at [-V] [-q x] [-f file] [-m] -t [[CC]YY]MMDDhhmm
      at -c job [job...]
      (atq | at -l) [-V] [-q x]
      (atrm | at -d | at -r) [-V] [-q x] job ...
      batch [-V] [-f file] [-m]
Usage: at [-V] [-q x] [-f file] [-m] time
      at [-V] [-q x] [-f file] [-m] -t [[CC]YY]MMDDhhmm
      at -c job [job...]
      (atq | at -l) [-V] [-q x]
      (atrm | at -d | at -r) [-V] [-q x] job ...
      batch [-V] [-f file] [-m]
   while executing
"exec batch -q long.q ${filebase}_script.sh"
   ("probtrackx" arm line 134)
   invoked from within
"switch -- $probtrack(tool) {
   eddy_current {
       global eddy

       set errorStr ""
       if { $eddy(input) == "" } { set errorStr "You need to specify..."
   (procedure "fdt:apply" line 5)
   invoked from within
"fdt:apply .fdt keep"
   invoked from within
".fdt.apply invoke"
   ("uplevel" body line 1)
   invoked from within
"uplevel #0 [list $w invoke]"
   (procedure "tk::ButtonUp" line 22)
   invoked from within
"tk::ButtonUp .fdt.apply"
   (command bound to event)


Running probtrackx from the command line with only the default options present gives the following error:
Log directory is: ++
7
sl 0
sl 1
sl 2
sl 3
sl 4
and so on... to sl 38 followed by
run
53 53 38
Segmentation fault

Could anyone help point us in the right direction as to what may be the problem ?

Thanks.

Mahinda

---------------------------------------------------------------------------
Saad Jbabdi, 
Postdoctoral Research Assistant,  
Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222545   (fax 222717)
---------------------------------------------------------------------------





------=_Part_16077_12249417.1192106681895-- ========================================================================= Date: Thu, 11 Oct 2007 09:14:17 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Zhang, Xiaochu (NIH/NIDA) [F]" <[log in to unmask]> Subject: hi and for VBM problem. MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01C80C08.A0C4E7CF" This is a multi-part message in MIME format. ------_=_NextPart_001_01C80C08.A0C4E7CF Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: quoted-printable Hi, I'm working on the VBM data. My sample included 48 subjects (24 patients and 24 controls). I have two questions about the "fslvbm_3_proc". The first one, I ran it 2 days ago. However, up to now, it have NOT finished. I checked the program is not died. However, when I run "fslvbm_1" and "fslvbm_2", it took no more than 4 hours. Is it normal phenomenon? The second one, about 1 day ago, my computer was rebooted by power problem. After reboot, I re-run the "fslvbm_3". It looks like OK. Need I to deleted all files and restart from "fslvbm_1"? =20 Thank you very much! Xiaochu Zhang Ph.D Visiting Research Fellow NIH/NIDA-IRP 5500 Nathan Shock Drive Baltimore MD 21224 =20 Tel: (410) - 550 - 1440 ext. 434 =20 ------_=_NextPart_001_01C80C08.A0C4E7CF Content-Type: text/html; charset="US-ASCII" Content-Transfer-Encoding: quoted-printable

Hi,

I’m working on the VBM data. My sample included = 48 subjects (24 patients and 24 controls). I have two questions about the = “fslvbm_3_proc”.

The first one, I ran it 2 days ago. However, up to = now, it have NOT finished. I checked the program is not died. However, when I = run “fslvbm_1” and “fslvbm_2”, it took no more than 4 hours. Is it normal phenomenon?

The second one, about 1 day ago, my computer was = rebooted by power problem. After reboot, I re-run the “fslvbm_3”. It = looks like OK. Need I to deleted all files and restart from “fslvbm_1”? =  

Thank you very much!

Xiaochu Zhang Ph.D

Visiting Research Fellow

NIH/NIDA-IRP

5500 Nathan Shock Drive

Baltimore<= font size=3D2 color=3Dnavy> = MD 21224

 

Tel: (410) - 550 - 1440 ext. 434

 

------_=_NextPart_001_01C80C08.A0C4E7CF-- ========================================================================= Date: Thu, 11 Oct 2007 15:38:59 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: =?iso-8859-1?q?Gwena=EBlle=20DOUAUD?= <[log in to unmask]> Subject: RE : [FSL] hi and for VBM problem. In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Hi, I'm not sure about your first question. The non-linear registration can take a while, let's say at least 2 hours per subject.=20 We have done the things so that you can run the jobs in parallel if your lab has got a cluster to do so. If this is not the case, yes, it can be long, and much longer than fslvbm2 for which you have hopefully chosen the recommended affine option with -a. Just check that your 4D image GM_mod_merg.nii.gz in your stats directory has been created by the most recent fslvbm3 you ran, and if so, just abort the first fslvbm3 command.=20 Hope this helps, Gwenaelle --- "Zhang, Xiaochu (NIH/NIDA) [F]" <[log in to unmask]> a =E9crit : > Hi, >=20 > I'm working on the VBM data. My sample included 48 > subjects (24 patients > and 24 controls). I have two questions about the > "fslvbm_3_proc". >=20 > The first one, I ran it 2 days ago. However, up to > now, it have NOT > finished. I checked the program is not died. > However, when I run > "fslvbm_1" and "fslvbm_2", it took no more than 4 > hours. Is it normal > phenomenon? >=20 > The second one, about 1 day ago, my computer was > rebooted by power > problem. After reboot, I re-run the "fslvbm_3". It > looks like OK. Need I > to deleted all files and restart from "fslvbm_1"? =20 >=20 > Thank you very much! >=20 > Xiaochu Zhang Ph.D >=20 > Visiting Research Fellow >=20 > NIH/NIDA-IRP >=20 > 5500 Nathan Shock Drive >=20 > Baltimore MD 21224 >=20 > =20 >=20 > Tel: (410) - 550 - 1440 ext. 434 >=20 > =20 >=20 >=20 ___________________________________________________________________= __________=20 Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! M= ail=20 ========================================================================= Date: Thu, 11 Oct 2007 14:47:53 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Dave Flitney <[log in to unmask]> Subject: Re: fslview: segmentation fault Comments: cc: Michael Hanke <[log in to unmask]> In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: multipart/alternative; boundary=Apple-Mail-2--945090972 --Apple-Mail-2--945090972 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Naj, I don't understand. The version referred to on the web page I sent you is clearly the latest version with the new online help and the atlas viewing support - both brand new in 3.x. When you start it up which version number is displayed? Michael, am I making some obvious mistake here? On 10 Oct 2007, at 18:52, Najmeh Khalili M. wrote: > Hi Dave, > > It seems the debian package you pointed to doesn't have the > recent fslview you suggested. Our sys admin is asking if you > plan to make a more recent or the most recent version of > flsview for the Debian "etch" distribution. > > We've tested the current available version of fslview > for etch, first by installing the binary distribution and then > by installing/building from source using apt-src. > > Thanks, > Naj > > > > > On Wed, 10 Oct 2007, Dave Flitney wrote: > >> Naj, >> >> I've switched maintenance efforts to Version 3.0.x now and, unless >> you have a really important reason for not using the new version, I'd >> ask that you upgrade FSLView. Debian packages are available, see >> http://apsy.gse.uni-magdeburg.de/fsl for details. >> >> On 8 Oct 2007, at 18:06, Najmeh Khalili M. wrote: >> >>> Hi Dave, >>> >>> I have the fslview (Version 2.4pre0), running on: >>> Debian GNU/Linux 4.0 \n \l >>> >>> rosaline:~# uname -a >>> Linux rosaline 2.6.22.2-i686-smp #1 SMP Tue Sep 25 14:47:07 EDT >>> 2007 i686 GNU/Linux >>> >>> Thanks, >>> Naj >>> >>> >>> >>> On Mon, 8 Oct 2007, Dave Flitney wrote: >>> >>>> Naj, >>>> >>>> It opens fine on the platforms I've tried so far - FSLView 3.0 on: >>>> MacOSX (PPC and Intel); Fedora Core 5 (64bit); and Centos4 (32bit). >>>> >>>> Which OS & FSLView versions are you using? >>>> >>>> On 5 Oct 2007, at 16:51, Najmeh Khalili M. wrote: >>>> >>>>> Hi Dave, >>>>> >>>>> Ref number is: >>>>> 943835 >>>>> >>>>> This is supposed to be a mask. It is not empty at fslstats >>>>> indicates non-zero content. >>>> >>>> Didn't think it was. That was a different poster :-) >>>> >>>>> Thanks >>>>> Naj >>>> >>>> -- >>>> Cheers, Dave >>>> >>>> Dave Flitney, IT Manager >>>> Oxford Centre for Functional MRI of the Brain >>>> E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 >>>> URL: http://www.fmrib.ox.ac.uk/~flitney >>>> >>>> >>>> >> >> -- >> Cheers, Dave >> >> Dave Flitney, IT Manager >> Oxford Centre for Functional MRI of the Brain >> E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 >> URL: http://www.fmrib.ox.ac.uk/~flitney >> >> >> -- Cheers, Dave Dave Flitney, IT Manager Oxford Centre for Functional MRI of the Brain E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 URL: http://www.fmrib.ox.ac.uk/~flitney --Apple-Mail-2--945090972 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1 Naj,

I don't understand. The = version referred to on the web page I sent you is clearly the latest = version with the new online help and the atlas viewing support - both = brand new in 3.x. When you start it up which version number is = displayed?

Michael, = am I making some obvious mistake here?

On 10 Oct = 2007, at 18:52, Najmeh Khalili M. wrote:

Hi Dave,

It seems the debian package you = pointed to doesn't have the
recent = fslview you suggested. Our sys admin is asking if you
plan to make a more recent or the most recent = version of
flsview for the Debian "etch" = distribution.

We've tested the current available version of = fslview
for etch, first by installing = the binary distribution and then
by = installing/building from source using apt-src.

Naj




On Wed, = 10 Oct 2007, Dave Flitney wrote:


I've switched maintenance efforts to Version 3.0.x = now and, unless
you have a really important = reason for not using the new version, I'd
ask that = you upgrade FSLView. Debian packages are available, see

On 8 Oct 2007, at 18:06, Najmeh = Khalili M. wrote:

=
Hi Dave,

I have = the=A0 fslview (Version = 2.4pre0), running on:
Debian GNU/Linux 4.0 \n = \l

rosaline:~# uname -a
Linux = rosaline 2.6.22.2-i686-smp #1 SMP Tue Sep 25 14:47:07 EDT
2007 i686 GNU/Linux

Naj



On Mon, 8 Oct 2007, Dave Flitney wrote:

Naj,

It opens = fine on the platforms I've tried so far - FSLView 3.0 on:
MacOSX (PPC and Intel); Fedora Core 5 (64bit); and = Centos4 (32bit).

Which OS & FSLView versions are you = using?

On 5 Oct 2007, at 16:51, Najmeh Khalili M. = wrote:

=
Hi Dave,

Ref = number is:
943835

This is = supposed to be a mask. It is not empty at fslstats
indicates non-zero content.

Didn't = think it was. That was a different poster :-)

Thanks
Naj

--
Cheers, Dave

Dave Flitney, IT = Manager
Oxford Centre for Functional MRI = of the Brain
E:[log in to unmask] = W:+44-1865-222713 F:+44-1865-222717




Cheers, Dave

Dave = Flitney, IT Manager
Oxford Centre for = Functional MRI of the Brain
E:[log in to unmask] = W:+44-1865-222713 F:+44-1865-222717




=
Dave Flitney, = IT Manager
Oxford Centre for Functional MRI = of the Brain
E:[log in to unmask] = W:+44-1865-222713 F:+44-1865-222717

=

= --Apple-Mail-2--945090972-- ========================================================================= Date: Thu, 11 Oct 2007 10:10:46 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Zhang, Xiaochu (NIH/NIDA) [F]" <[log in to unmask]> Subject: Re: RE : [FSL] hi and for VBM problem. In-Reply-To: A<[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: base64 VGhhbmsgeW91IHZlcnkgbXVjaCBmb3Igc3BlZWR5IHJlc3BvbnNlIQ0KSSB0aGluayB0aGUgZmly c3QgcXVlc3Rpb24gaGFzIGJlZW4gcmVzb2x2ZWQuIEZvciAyIGhvdXJzIHBlciBzdWJqZWN0LCBp dCB3aWxsIHRha2UgYWJvdXQgMTAwIGhvdXJzIChhYm91dCA0IGRheXMpIHRvIGZpbmlzaCB0aGUg d2hvbGUgNDggc3ViamVjdHMuIEkgbmVlZCB3YWl0IHR3byBkYXlzIG1vcmUuDQpGb3IgeW91IHNh aWQgaXQgY2FuIGJlIHJ1biBpbiBwYXJhbGxlbCBhbmQgdGhlIGNvbXB1dGVyIGluIG91ciBsYWIg aGF2ZSBtdWx0aXBsZSBDUFVzLiBTbywgY2FuIEkgb3BlbiBtdWx0aXBsZSB0ZXJtaW5hbHMgYW5k IHJ1biB0aGUgc2NyaXB0IGluIGVhY2ggb25lIHRvIHNob3J0ZW4gdGhlIHdob2xlIHRpbWU/DQpU aGFua3MgYWdhaW4hDQoNCi0tLS0tT3JpZ2luYWwgTWVzc2FnZS0tLS0tDQpGcm9tOiBHd2VuYcOr bGxlIERPVUFVRCBbbWFpbHRvOmd3ZW5hZWxsZWRvdWF1ZEBZQUhPTy5GUl0gDQpTZW50OiAyMDA3 5bm0MTDmnIgxMeaXpSA5OjM5DQpUbzogRlNMQEpJU0NNQUlMLkFDLlVLDQpTdWJqZWN0OiBbRlNM XSBSRSA6IFtGU0xdIGhpIGFuZCBmb3IgVkJNIHByb2JsZW0uDQoNCkhpLA0KDQpJJ20gbm90IHN1 cmUgYWJvdXQgeW91ciBmaXJzdCBxdWVzdGlvbi4gVGhlIG5vbi1saW5lYXINCnJlZ2lzdHJhdGlv biBjYW4gdGFrZSBhIHdoaWxlLCBsZXQncyBzYXkgYXQgbGVhc3QgMg0KaG91cnMgcGVyIHN1Ympl Y3QuIA0KV2UgaGF2ZSBkb25lIHRoZSB0aGluZ3Mgc28gdGhhdCB5b3UgY2FuIHJ1biB0aGUgam9i cw0KaW4gcGFyYWxsZWwgaWYgeW91ciBsYWIgaGFzIGdvdCBhIGNsdXN0ZXIgdG8gZG8gc28uIElm DQp0aGlzIGlzIG5vdCB0aGUgY2FzZSwgeWVzLCBpdCBjYW4gYmUgbG9uZywgYW5kIG11Y2gNCmxv bmdlciB0aGFuIGZzbHZibTIgZm9yIHdoaWNoIHlvdSBoYXZlIGhvcGVmdWxseQ0KY2hvc2VuIHRo ZSByZWNvbW1lbmRlZCBhZmZpbmUgb3B0aW9uIHdpdGggLWEuDQpKdXN0IGNoZWNrIHRoYXQgeW91 ciA0RCBpbWFnZSBHTV9tb2RfbWVyZy5uaWkuZ3ogaW4NCnlvdXIgc3RhdHMgZGlyZWN0b3J5IGhh cyBiZWVuIGNyZWF0ZWQgYnkgdGhlIG1vc3QNCnJlY2VudCBmc2x2Ym0zIHlvdSByYW4sIGFuZCBp ZiBzbywganVzdCBhYm9ydCB0aGUNCmZpcnN0ICBmc2x2Ym0zIGNvbW1hbmQuIA0KDQpIb3BlIHRo aXMgaGVscHMsDQpHd2VuYWVsbGUNCg0KDQotLS0gIlpoYW5nLCBYaWFvY2h1IChOSUgvTklEQSkg W0ZdIg0KPHpoYW5neDJASU5UUkEuTklEQS5OSUguR09WPiBhIMOpY3JpdCA6DQoNCj4gSGksDQo+ IA0KPiBJJ20gd29ya2luZyBvbiB0aGUgVkJNIGRhdGEuIE15IHNhbXBsZSBpbmNsdWRlZCA0OA0K PiBzdWJqZWN0cyAoMjQgcGF0aWVudHMNCj4gYW5kIDI0IGNvbnRyb2xzKS4gSSBoYXZlIHR3byBx dWVzdGlvbnMgYWJvdXQgdGhlDQo+ICJmc2x2Ym1fM19wcm9jIi4NCj4gDQo+IFRoZSBmaXJzdCBv bmUsIEkgcmFuIGl0IDIgZGF5cyBhZ28uIEhvd2V2ZXIsIHVwIHRvDQo+IG5vdywgaXQgaGF2ZSBO T1QNCj4gZmluaXNoZWQuIEkgY2hlY2tlZCB0aGUgcHJvZ3JhbSBpcyBub3QgZGllZC4NCj4gSG93 ZXZlciwgd2hlbiBJIHJ1bg0KPiAiZnNsdmJtXzEiIGFuZCAiZnNsdmJtXzIiLCBpdCB0b29rIG5v IG1vcmUgdGhhbiA0DQo+IGhvdXJzLiBJcyBpdCBub3JtYWwNCj4gcGhlbm9tZW5vbj8NCj4gDQo+ IFRoZSBzZWNvbmQgb25lLCBhYm91dCAxIGRheSBhZ28sIG15IGNvbXB1dGVyIHdhcw0KPiByZWJv b3RlZCBieSBwb3dlcg0KPiBwcm9ibGVtLiBBZnRlciByZWJvb3QsIEkgcmUtcnVuIHRoZSAiZnNs dmJtXzMiLiBJdA0KPiBsb29rcyBsaWtlIE9LLiBOZWVkIEkNCj4gdG8gZGVsZXRlZCBhbGwgZmls ZXMgYW5kIHJlc3RhcnQgZnJvbSAiZnNsdmJtXzEiPyAgDQo+IA0KPiBUaGFuayB5b3UgdmVyeSBt dWNoIQ0KPiANCj4gWGlhb2NodSBaaGFuZyBQaC5EDQo+IA0KPiBWaXNpdGluZyBSZXNlYXJjaCBG ZWxsb3cNCj4gDQo+IE5JSC9OSURBLUlSUA0KPiANCj4gNTUwMCBOYXRoYW4gU2hvY2sgRHJpdmUN Cj4gDQo+IEJhbHRpbW9yZSBNRCAyMTIyNA0KPiANCj4gIA0KPiANCj4gVGVsOiAoNDEwKSAtIDU1 MCAtIDE0NDAgZXh0LiA0MzQNCj4gDQo+ICANCj4gDQo+IA0KDQoNCg0KICAgICAgX19fX19fX19f X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19f X19fX19fX19fX18gDQpOZSBnYXJkZXogcGx1cyBxdSd1bmUgc2V1bGUgYWRyZXNzZSBtYWlsICEg Q29waWV6IHZvcyBtYWlscyB2ZXJzIFlhaG9vISBNYWlsIA0K ========================================================================= Date: Thu, 11 Oct 2007 16:19:22 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Michael Hanke <[log in to unmask]> Subject: Re: fslview: segmentation fault Comments: To: Dave Flitney <[log in to unmask]> In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Hi, On Thu, Oct 11, 2007 at 02:47:53PM +0100, Dave Flitney wrote: > Naj, > I don't understand. The version referred to on the web page I sent you is > clearly the latest version with the new online help and the atlas viewing > support - both brand new in 3.x. When you start it up which version number > is displayed? > Michael, am I making some obvious mistake here? No, everything is correct. The latest version of fslview for Debian etch is: 3.0+4.0.1-2. Meaning fslview 3.0 from FSL 4.0.1 package version 2. I'm using this package on etch right now, so I'm pretty confident that it works. Naj, please make sure that you actually use the package named 'fslview' not the one called 'fsl-fslview'. The latter one is discontinued and should be automatically replaced by 'fslview'. 'fsl-fslview' is not available anymore, but still might be installed on your machine. If you cannot find this package in you package list, you might have to update your package cache (e.g. aptitude update). Hope that helps to clarify things a bit. Cheers, Michael -- GPG key: 1024D/3144BE0F Michael Hanke http://apsy.gse.uni-magdeburg.de/hanke ICQ: 48230050 ========================================================================= Date: Thu, 11 Oct 2007 16:35:38 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: =?iso-8859-1?q?Gwena=EBlle=20DOUAUD?= <[log in to unmask]> Subject: RE : Re: [FSL] RE : [FSL] hi and for VBM problem. In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Hi, sorry, I should have been more specific... It can be run in parallel if your lab has got a SGE system or something like that (sorry I'm not very good at these things) and then you send your job to a bunch of computers involved in the cluster. It should not take much longer anyway.=20 Hopefully, we should get a better and much quicker non-linear registration (FNIRT) soonish and that will be included in the next FSL4.1 patch. Cheers, Gwenaelle =20 --- "Zhang, Xiaochu (NIH/NIDA) [F]" <[log in to unmask]> a =E9crit : > Thank you very much for speedy response! > I think the first question has been resolved. For 2 > hours per subject, it will take about 100 hours > (about 4 days) to finish the whole 48 subjects. I > need wait two days more. > For you said it can be run in parallel and the > computer in our lab have multiple CPUs. So, can I > open multiple terminals and run the script in each > one to shorten the whole time? > Thanks again! >=20 > -----Original Message----- > From: Gwena=C3=ABlle DOUAUD > [mailto:[log in to unmask]]=20 > Sent: 2007=E5=B9=B410=E6=9C=8811=E6=97=A5 9:39 > To: [log in to unmask] > Subject: [FSL] RE : [FSL] hi and for VBM problem. >=20 > Hi, >=20 > I'm not sure about your first question. The > non-linear > registration can take a while, let's say at least 2 > hours per subject.=20 > We have done the things so that you can run the jobs > in parallel if your lab has got a cluster to do so. > If > this is not the case, yes, it can be long, and much > longer than fslvbm2 for which you have hopefully > chosen the recommended affine option with -a. > Just check that your 4D image GM_mod_merg.nii.gz in > your stats directory has been created by the most > recent fslvbm3 you ran, and if so, just abort the > first fslvbm3 command.=20 >=20 > Hope this helps, > Gwenaelle >=20 >=20 > --- "Zhang, Xiaochu (NIH/NIDA) [F]" > <[log in to unmask]> a =C3=A9crit : >=20 > > Hi, > >=20 > > I'm working on the VBM data. My sample included 48 > > subjects (24 patients > > and 24 controls). I have two questions about the > > "fslvbm_3_proc". > >=20 > > The first one, I ran it 2 days ago. However, up to > > now, it have NOT > > finished. I checked the program is not died. > > However, when I run > > "fslvbm_1" and "fslvbm_2", it took no more than 4 > > hours. Is it normal > > phenomenon? > >=20 > > The second one, about 1 day ago, my computer was > > rebooted by power > > problem. After reboot, I re-run the "fslvbm_3". It > > looks like OK. Need I > > to deleted all files and restart from "fslvbm_1"?=20 >=20 > >=20 > > Thank you very much! > >=20 > > Xiaochu Zhang Ph.D > >=20 > > Visiting Research Fellow > >=20 > > NIH/NIDA-IRP > >=20 > > 5500 Nathan Shock Drive > >=20 > > Baltimore MD 21224 > >=20 > > =20 > >=20 > > Tel: (410) - 550 - 1440 ext. 434 > >=20 > > =20 > >=20 > >=20 >=20 >=20 >=20 > =20 > _________________________________________________________________________= ____ >=20 > Ne gardez plus qu'une seule adresse mail ! Copiez > vos mails vers Yahoo! Mail=20 >=20 ___________________________________________________________________= __________=20 Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! M= ail=20 ========================================================================= Date: Thu, 11 Oct 2007 10:39:23 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Najmeh Khalili M." <[log in to unmask]> Subject: Re: fslview: segmentation fault Comments: cc: Sylvain MILOT <[log in to unmask]> In-Reply-To: <20071011141922.GA22772@runkel> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi all, The fslview that I am using is the fslview3.0 (according to the title displayed.) I think the confusion about version came from doing fslview -h, which produces: ======================================= fslview (Version 2.4pre0) Copyright(c) 2005, University of Oxford Dave Flitney Usage: fslview [-m 3d|ortho|lightbox] [-l lutname] [-b low,hi] [ [-l lutname] [-b low,hi] ] ... ======================================== So perhaps the help file is not updated?! There is only one fslview installed on my computer. And that one still failes on the file I sent earlier, even after the upgrade. Now I am even more confused :) Cheers Naj On Thu, 11 Oct 2007, Michael Hanke wrote: > Hi, > > On Thu, Oct 11, 2007 at 02:47:53PM +0100, Dave Flitney wrote: > > Naj, > > I don't understand. The version referred to on the web page I sent you is > > clearly the latest version with the new online help and the atlas viewing > > support - both brand new in 3.x. When you start it up which version number > > is displayed? > > Michael, am I making some obvious mistake here? > No, everything is correct. The latest version of fslview for Debian etch > is: 3.0+4.0.1-2. Meaning fslview 3.0 from FSL 4.0.1 package version 2. > I'm using this package on etch right now, so I'm pretty confident that > it works. > > Naj, please make sure that you actually use the package named 'fslview' > not the one called 'fsl-fslview'. The latter one is discontinued and > should be automatically replaced by 'fslview'. 'fsl-fslview' is not > available anymore, but still might be installed on your machine. > > If you cannot find this package in you package list, you might have to > update your package cache (e.g. aptitude update). > > Hope that helps to clarify things a bit. > > Cheers, > > Michael > > > -- > GPG key: 1024D/3144BE0F Michael Hanke > http://apsy.gse.uni-magdeburg.de/hanke > ICQ: 48230050 > ========================================================================= Date: Thu, 11 Oct 2007 10:51:02 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matt Glasser <[log in to unmask]> Organization: ma-tea.com Subject: Re: RE : Re: [FSL] RE : [FSL] hi and for VBM problem. In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Gwenaelle, Will this non-linear registration algorithm replace ITK in TBSS? =20 Thanks, Matt. -----Original Message----- From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On = Behalf Of Gwena=C3=ABlle DOUAUD Sent: Thursday, October 11, 2007 10:36 AM To: [log in to unmask] Subject: [FSL] RE : Re: [FSL] RE : [FSL] hi and for VBM problem. Hi, sorry, I should have been more specific... It can be run in parallel if your lab has got a SGE system or something like that (sorry I'm not very good at these things) and then you send your job to a bunch of computers involved in the cluster. It should not take much longer anyway.=20 Hopefully, we should get a better and much quicker non-linear registration (FNIRT) soonish and that will be included in the next FSL4.1 patch. Cheers, Gwenaelle =20 --- "Zhang, Xiaochu (NIH/NIDA) [F]" <[log in to unmask]> a =C3=A9crit : > Thank you very much for speedy response! > I think the first question has been resolved. For 2 > hours per subject, it will take about 100 hours > (about 4 days) to finish the whole 48 subjects. I > need wait two days more. > For you said it can be run in parallel and the > computer in our lab have multiple CPUs. So, can I > open multiple terminals and run the script in each > one to shorten the whole time? > Thanks again! >=20 > -----Original Message----- > From: Gwena=C3=83=C2=ABlle DOUAUD > [mailto:[log in to unmask]]=20 > Sent: = 2007=C3=A5=C2=B9=C2=B410=C3=A6=C5=93=CB=8611=C3=A6=E2=80=94=C2=A5 9:39 > To: [log in to unmask] > Subject: [FSL] RE : [FSL] hi and for VBM problem. >=20 > Hi, >=20 > I'm not sure about your first question. The > non-linear > registration can take a while, let's say at least 2 > hours per subject.=20 > We have done the things so that you can run the jobs > in parallel if your lab has got a cluster to do so. > If > this is not the case, yes, it can be long, and much > longer than fslvbm2 for which you have hopefully > chosen the recommended affine option with -a. > Just check that your 4D image GM_mod_merg.nii.gz in > your stats directory has been created by the most > recent fslvbm3 you ran, and if so, just abort the > first fslvbm3 command.=20 >=20 > Hope this helps, > Gwenaelle >=20 >=20 > --- "Zhang, Xiaochu (NIH/NIDA) [F]" > <[log in to unmask]> a =C3=83=C2=A9crit : >=20 > > Hi, > >=20 > > I'm working on the VBM data. My sample included 48 > > subjects (24 patients > > and 24 controls). I have two questions about the > > "fslvbm_3_proc". > >=20 > > The first one, I ran it 2 days ago. However, up to > > now, it have NOT > > finished. I checked the program is not died. > > However, when I run > > "fslvbm_1" and "fslvbm_2", it took no more than 4 > > hours. Is it normal > > phenomenon? > >=20 > > The second one, about 1 day ago, my computer was > > rebooted by power > > problem. After reboot, I re-run the "fslvbm_3". It > > looks like OK. Need I > > to deleted all files and restart from "fslvbm_1"?=20 >=20 > >=20 > > Thank you very much! > >=20 > > Xiaochu Zhang Ph.D > >=20 > > Visiting Research Fellow > >=20 > > NIH/NIDA-IRP > >=20 > > 5500 Nathan Shock Drive > >=20 > > Baltimore MD 21224 > >=20 > > =20 > >=20 > > Tel: (410) - 550 - 1440 ext. 434 > >=20 > > =20 > >=20 > >=20 >=20 >=20 >=20 > =20 > _________________________________________________________________________= ____ >=20 > Ne gardez plus qu'une seule adresse mail ! Copiez > vos mails vers Yahoo! Mail=20 >=20 = _________________________________________________________________________= ____=20 Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! = Mail=20 ========================================================================= Date: Thu, 11 Oct 2007 16:55:01 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: =?iso-8859-1?q?Gwena=EBlle=20DOUAUD?= <[log in to unmask]> Subject: RE : Re: [FSL] RE : Re: [FSL] RE : [FSL] hi and for VBM problem. In-Reply-To: [log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Hi, as far as I know, yes, absolutely... Cheers, Gwenaelle --- Matt Glasser <[log in to unmask]> a =E9crit : > Gwenaelle, >=20 > Will this non-linear registration algorithm replace > ITK in TBSS? =20 >=20 > Thanks, >=20 > Matt. >=20 > -----Original Message----- > From: FSL - FMRIB's Software Library > [mailto:[log in to unmask]] On Behalf Of Gwena=C3=ABlle > DOUAUD > Sent: Thursday, October 11, 2007 10:36 AM > To: [log in to unmask] > Subject: [FSL] RE : Re: [FSL] RE : [FSL] hi and for > VBM problem. >=20 > Hi, >=20 > sorry, I should have been more specific... It can be > run in parallel if your lab has got a SGE system or > something like that (sorry I'm not very good at > these > things) and then you send your job to a bunch of > computers involved in the cluster. It should not > take > much longer anyway.=20 > Hopefully, we should get a better and much quicker > non-linear registration (FNIRT) soonish and that > will > be included in the next FSL4.1 patch. >=20 > Cheers, > Gwenaelle >=20 > =20 > --- "Zhang, Xiaochu (NIH/NIDA) [F]" > <[log in to unmask]> a =C3=A9crit : >=20 > > Thank you very much for speedy response! > > I think the first question has been resolved. For > 2 > > hours per subject, it will take about 100 hours > > (about 4 days) to finish the whole 48 subjects. I > > need wait two days more. > > For you said it can be run in parallel and the > > computer in our lab have multiple CPUs. So, can I > > open multiple terminals and run the script in each > > one to shorten the whole time? > > Thanks again! > >=20 > > -----Original Message----- > > From: Gwena=C3=83=C2=ABlle DOUAUD > > [mailto:[log in to unmask]]=20 > > Sent: 2007=C3=A5=C2=B9=C2=B410=C3=A6=C5=93=CB=8611=C3=A6=E2=80=94=C2=A5= 9:39 > > To: [log in to unmask] > > Subject: [FSL] RE : [FSL] hi and for VBM problem. > >=20 > > Hi, > >=20 > > I'm not sure about your first question. The > > non-linear > > registration can take a while, let's say at least > 2 > > hours per subject.=20 > > We have done the things so that you can run the > jobs > > in parallel if your lab has got a cluster to do > so. > > If > > this is not the case, yes, it can be long, and > much > > longer than fslvbm2 for which you have hopefully > > chosen the recommended affine option with -a. > > Just check that your 4D image GM_mod_merg.nii.gz > in > > your stats directory has been created by the most > > recent fslvbm3 you ran, and if so, just abort the > > first fslvbm3 command.=20 > >=20 > > Hope this helps, > > Gwenaelle > >=20 > >=20 > > --- "Zhang, Xiaochu (NIH/NIDA) [F]" > > <[log in to unmask]> a =C3=83=C2=A9crit : > >=20 > > > Hi, > > >=20 > > > I'm working on the VBM data. My sample included > 48 > > > subjects (24 patients > > > and 24 controls). I have two questions about the > > > "fslvbm_3_proc". > > >=20 > > > The first one, I ran it 2 days ago. However, up > to > > > now, it have NOT > > > finished. I checked the program is not died. > > > However, when I run > > > "fslvbm_1" and "fslvbm_2", it took no more than > 4 > > > hours. Is it normal > > > phenomenon? > > >=20 > > > The second one, about 1 day ago, my computer was > > > rebooted by power > > > problem. After reboot, I re-run the "fslvbm_3". > It > > > looks like OK. Need I > > > to deleted all files and restart from > "fslvbm_1"?=20 > >=20 > > >=20 > > > Thank you very much! > > >=20 > > > Xiaochu Zhang Ph.D > > >=20 > > > Visiting Research Fellow > > >=20 > > > NIH/NIDA-IRP > > >=20 > > > 5500 Nathan Shock Drive > > >=20 > > > Baltimore MD 21224 > > >=20 > > > =20 > > >=20 > > > Tel: (410) - 550 - 1440 ext. 434 > > >=20 > > > =20 > > >=20 > > >=20 > >=20 > >=20 > >=20 > > =20 > > > _________________________________________________________________________= ____ > >=20 > > Ne gardez plus qu'une seule adresse mail ! Copiez > > vos mails vers Yahoo! Mail=20 > >=20 >=20 >=20 >=20 > =20 > _________________________________________________________________________= ____ >=20 > Ne gardez plus qu'une seule adresse mail ! Copiez > vos mails vers Yahoo! Mail=20 >=20 ___________________________________________________________________= __________=20 Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! M= ail=20 ========================================================================= Date: Thu, 11 Oct 2007 11:09:02 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Daniel Glen <[log in to unmask]> Subject: Re: AW: [FSL] Blood Flow In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: multipart/alternative; boundary=Apple-Mail-2--940221742 --Apple-Mail-2--940221742 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed The value, first determined by Seymour Kety and Carl Schmidt in 1948 using an overall brain measurement found the value to be 54 ml/100g (tissue)/min. Later techniques were developed to determine local blood flow in humans using inert radioisotopic gases such as 85-Kr and then 133-Xe found values ranging from 10 to 155 ml/g/min (Ingvar and Lassen, 1961, 1985). Measurements using H15-2-0 water by Herscovitch and Raichle were then done in PET in 1985. Overall flow values have ranged from 39 to 54 ml/100g/min depending upon the methods and subject population. There was this 2000 paper that includes a discussion of various methods including the newer sonographic method They report a range of values from 567-668 ml/min for overall flow and 48.5 ml/100g/min. http://www.nature.com/jcbfm/journal/v20/n2/full/9590879a.html Regards, Daniel Glen On Oct 11, 2007, at 7:36 AM, M. Pham wrote: > Hi, > > an often quoted normal human CBF value is: > > ~ 50ml / 100ml /min (to my knowledge this value is not very > different in animals relevant to experimental cerebral ischemia) > > CBF landmarks in ischemia mostly derived from animal models are: > > ~ 10-20ml ==> ischemia associated with increasingly abnormal > neuronal function (often determined by EEG) and biochemical > abnormalities (acidosis/ATP depletion) - at this level of ischemia > alterations are reversible if blood flow is restored timely > (concept of ischemic penumbra/corona) > > < 10 ml ==> rapidly ensuing infarction/cell death regardless of the > duration of ischemia > > Hope this helps! > Mirko > > Von: FSL - FMRIB's Software Library [mailto:[log in to unmask]] Im > Auftrag von cheng ouyang > Gesendet: Donnerstag, 11. Oktober 2007 00:21 > An: [log in to unmask] > Betreff: [FSL] Blood Flow > > Hi, FSLers, > > I have a question about blood flow of human brain. Normally, people > like to use CBF (ml/100ml/min) to describe the blood flow. My > question is that Anybody know about the approximate numbers of > BLODD FlOW RATE(ml/min) in human cortex (maybe capillary, arteriole)? > > > Cathy > > > Take the Internet to Go: Yahoo!Go puts the Internet in your pocket: > mail, news, photos & more. --Apple-Mail-2--940221742 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1 The value, first determined by = Seymour Kety and Carl Schmidt in 1948 using an overall brain measurement = found the value to be 54 ml/100g (tissue)/min. Later techniques were = developed to determine local blood flow in humans using inert = radioisotopic gases such as 85-Kr and then 133-Xe found values ranging = from 10 to 155 ml/g/min (Ingvar and Lassen, 1961, 1985).=A0 Measurements = using H15-2-0 water by Herscovitch and Raichle were then done in PET in = 1985. Overall flow values have ranged from 39 to 54 ml/100g/min = depending upon the methods and subject population. There was this 2000 = paper that includes a discussion of various methods including the newer = sonographic method They report a range of values from 567-668 ml/min for = overall flow and 48.5 ml/100g/min.

htt= p://www.nature.com/jcbfm/journal/v20/n2/full/9590879a.html


Regards,
Daniel = Glen


On Oct 11, = 2007, at 7:36 AM, M. Pham wrote:

Hi,
=
=A0
an often quoted normal human = CBF value is:
=A0
~=A050ml / 100ml /min=A0(to my knowledge this value is not=A0very different in animals = relevant to experimental cerebral ischemia)
=A0
CBF landmarks in ischemia mostly derived from animal models = are:
=A0
~ 10-20ml =3D=3D> ischemia associated with increasingly = abnormal neuronal function (often determined by EEG) and biochemical = abnormalities (acidosis/ATP depletion) - at this level of ischemia = alterations are=A0reversible if blood flow is restored timely=A0(concept = of ischemic penumbra/corona)
=A0
< 10 ml =3D=3D> rapidly = ensuing=A0infarction/cell death regardless of the duration of = ischemia
=A0
Hope this helps!
Mirko

=
Von: FSL - = FMRIB's Software Library [mailto:[log in to unmask]] Im Auftrag von = cheng ouyang
Gesendet: Donnerstag, 11. Oktober 2007 = 00:21
An: [log in to unmask]
Betreff: [FSL] Blood = Flow

Hi, FSLers,

I have a = question about blood flow of human brain. Normally, people like to use = CBF (ml/100ml/min) to describe the blood flow. My question is that = Anybody know about the approximate numbers of BLODD FlOW RATE(ml/min) in = human cortex (maybe capillary, arteriole)?


Cathy

=

Take the = Internet to Go: Yahoo!Go puts the Internet in your pocket: mail, news, photos & more. =

= --Apple-Mail-2--940221742-- ========================================================================= Date: Thu, 11 Oct 2007 16:30:12 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Christopher Petty <[log in to unmask]> Subject: Re: first_utils Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="windows-1252" sorry to keep piling up the questions, but if i set up my regression for=20= first_utils as one group, with one regressor, then only try to look at=20= that regressor (just like i would in feat), where does first_utils split=20= them to create the 2 means displayed in the .vtk? ie: EV1 EV2 1 10 1 5 1 30 1 12 contrast: 0 1 ========================================================================= Date: Thu, 11 Oct 2007 11:39:22 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Ted Yanagihara <[log in to unmask]> Subject: running read_avw commands in matlab MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_17158_19832305.1192117162515" ------=_Part_17158_19832305.1192117162515 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hello list, I am trying to read some of my images from FEAT and probtrackx in matlab. I have been getting an error when using read_avw and read_avw_complex, but the other commands (such as read_avw_hdr) work fine. After looking at the scripts for these functions, it appears that matlab cannot find the path for fsl.sh because it wants to use the script says to use this path: ${FSLDIR}/etc/fslconf/fsl.sh but matlab reads this as: /etc/fslconf/fsl.sh Is this problem because matlab is running from bash as its default when I need it to run in tcsh? If so, does anyone know how I can make my default environment tcsh? If not, what else might be going on? Here is the full error that matlab gives me when I run this set of commands: >> dir='~/Lab/Data/AKL_13/dti_FA'; >> fid=fopen(dir); >> image=read_avw(fid) Warning: The argument for the %s format specifier must be of type char (a string). > In read_avw at 30 sh: line 1: /etc/fslconf/fsl.sh: No such file or directory ??? Error using ==> fread Invalid file identifier. Use fopen to generate a valid file identifier. Error in ==> read_avw_hdr at 17 testval = fread(fid,1,'int32'); Error in ==> read_avw at 33 [dims,scales,bpp,endian,datatype]= read_avw_hdr(tmpname); ------=_Part_17158_19832305.1192117162515 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hello list,

I am trying to read some of my images from FEAT and probtrackx in matlab. I have been getting an error when using read_avw and read_avw_complex, but the other commands (such as read_avw_hdr) work fine.

After looking at the scripts for these functions, it appears that matlab cannot find the path for fsl.sh because it wants to use the script says to use this path: ${FSLDIR}/etc/fslconf/fsl.sh   but matlab reads this as: /etc/fslconf/fsl.sh

Is this problem because matlab is running from bash as its default when I need it to run in tcsh? If so, does anyone know how I can make my default environment tcsh? If not, what else might be going on?


Here is the full error that matlab gives me when I run this set of commands:

>> dir='~/Lab/Data/AKL_13/dti_FA';
>> fid=fopen(dir);
>> image=read_avw(fid)
Warning: The argument for the %s format specifier must be of type char (a string).
> In read_avw at 30
sh: line 1: /etc/fslconf/fsl.sh: No such file or directory
??? Error using ==> fread
Invalid file identifier.  Use fopen to generate a valid file identifier.

Error in ==> read_avw_hdr at 17
testval = fread(fid,1,'int32');

Error in ==> read_avw at 33
  [dims,scales,bpp,endian,datatype]= read_avw_hdr(tmpname);
------=_Part_17158_19832305.1192117162515-- ========================================================================= Date: Thu, 11 Oct 2007 12:02:01 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Zhang, Xiaochu (NIH/NIDA) [F]" <[log in to unmask]> Subject: Re: RE : Re: [FSL] RE : [FSL] hi and for VBM problem. In-Reply-To: A<[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: base64 VGhhbmsgeW91IHZlcnkgbXVjaCEgDQpTbyBteSBxdWVzdGlvbiBpcyB3aGV0aGVyIEkgYW0gYWJs ZSB0byBpbnRlcnJ1cHQgdGhlIHByb2dyYW0gd2hlbiBpdCBpcyBydW5uaW5nIChGb3IgaXQgdG9v ayBhIGxvbmcgdGltZSwgc29tZXRpbWVzIGl0IGNhbiBub3QgYmUgYXZvaWQuICkgYW5kIHJlc3Rh cnQgaXQgYWdhaW4uDQpEb2VzIGl0IGluZmx1ZW5jZSB0aGUgZmluYWwgcmVzdWx0Pw0KVGhhbmtz LCBhZ2FpbiENCg0KLS0tLS1PcmlnaW5hbCBNZXNzYWdlLS0tLS0NCkZyb206IEd3ZW5hw6tsbGUg RE9VQVVEIFttYWlsdG86Z3dlbmFlbGxlZG91YXVkQFlBSE9PLkZSXSANClNlbnQ6IDIwMDflubQx MOaciDEx5pelIDEwOjM2DQpUbzogRlNMQEpJU0NNQUlMLkFDLlVLDQpTdWJqZWN0OiBbRlNMXSBS RSA6IFJlOiBbRlNMXSBSRSA6IFtGU0xdIGhpIGFuZCBmb3IgVkJNIHByb2JsZW0uDQoNCkhpLA0K DQpzb3JyeSwgSSBzaG91bGQgaGF2ZSBiZWVuIG1vcmUgc3BlY2lmaWMuLi4gSXQgY2FuIGJlDQpy dW4gaW4gcGFyYWxsZWwgaWYgeW91ciBsYWIgaGFzIGdvdCBhIFNHRSBzeXN0ZW0gb3INCnNvbWV0 aGluZyBsaWtlIHRoYXQgKHNvcnJ5IEknbSBub3QgdmVyeSBnb29kIGF0IHRoZXNlDQp0aGluZ3Mp IGFuZCB0aGVuIHlvdSBzZW5kIHlvdXIgam9iIHRvIGEgYnVuY2ggb2YNCmNvbXB1dGVycyBpbnZv bHZlZCBpbiB0aGUgY2x1c3Rlci4gSXQgc2hvdWxkIG5vdCB0YWtlDQptdWNoIGxvbmdlciBhbnl3 YXkuIA0KSG9wZWZ1bGx5LCB3ZSBzaG91bGQgZ2V0IGEgYmV0dGVyIGFuZCBtdWNoIHF1aWNrZXIN Cm5vbi1saW5lYXIgcmVnaXN0cmF0aW9uIChGTklSVCkgc29vbmlzaCBhbmQgdGhhdCB3aWxsDQpi ZSBpbmNsdWRlZCBpbiB0aGUgbmV4dCBGU0w0LjEgcGF0Y2guDQoNCkNoZWVycywNCkd3ZW5hZWxs ZQ0KDQogDQotLS0gIlpoYW5nLCBYaWFvY2h1IChOSUgvTklEQSkgW0ZdIg0KPHpoYW5neDJASU5U UkEuTklEQS5OSUguR09WPiBhIMOpY3JpdCA6DQoNCj4gVGhhbmsgeW91IHZlcnkgbXVjaCBmb3Ig c3BlZWR5IHJlc3BvbnNlIQ0KPiBJIHRoaW5rIHRoZSBmaXJzdCBxdWVzdGlvbiBoYXMgYmVlbiBy ZXNvbHZlZC4gRm9yIDINCj4gaG91cnMgcGVyIHN1YmplY3QsIGl0IHdpbGwgdGFrZSBhYm91dCAx MDAgaG91cnMNCj4gKGFib3V0IDQgZGF5cykgdG8gZmluaXNoIHRoZSB3aG9sZSA0OCBzdWJqZWN0 cy4gSQ0KPiBuZWVkIHdhaXQgdHdvIGRheXMgbW9yZS4NCj4gRm9yIHlvdSBzYWlkIGl0IGNhbiBi ZSBydW4gaW4gcGFyYWxsZWwgYW5kIHRoZQ0KPiBjb21wdXRlciBpbiBvdXIgbGFiIGhhdmUgbXVs dGlwbGUgQ1BVcy4gU28sIGNhbiBJDQo+IG9wZW4gbXVsdGlwbGUgdGVybWluYWxzIGFuZCBydW4g dGhlIHNjcmlwdCBpbiBlYWNoDQo+IG9uZSB0byBzaG9ydGVuIHRoZSB3aG9sZSB0aW1lPw0KPiBU aGFua3MgYWdhaW4hDQo+IA0KPiAtLS0tLU9yaWdpbmFsIE1lc3NhZ2UtLS0tLQ0KPiBGcm9tOiBH d2VuYcODwqtsbGUgRE9VQVVEDQo+IFttYWlsdG86Z3dlbmFlbGxlZG91YXVkQFlBSE9PLkZSXSAN Cj4gU2VudDogMjAwN8OlwrnCtDEww6bFk8uGMTHDpuKAlMKlIDk6MzkNCj4gVG86IEZTTEBKSVND TUFJTC5BQy5VSw0KPiBTdWJqZWN0OiBbRlNMXSBSRSA6IFtGU0xdIGhpIGFuZCBmb3IgVkJNIHBy b2JsZW0uDQo+IA0KPiBIaSwNCj4gDQo+IEknbSBub3Qgc3VyZSBhYm91dCB5b3VyIGZpcnN0IHF1 ZXN0aW9uLiBUaGUNCj4gbm9uLWxpbmVhcg0KPiByZWdpc3RyYXRpb24gY2FuIHRha2UgYSB3aGls ZSwgbGV0J3Mgc2F5IGF0IGxlYXN0IDINCj4gaG91cnMgcGVyIHN1YmplY3QuIA0KPiBXZSBoYXZl IGRvbmUgdGhlIHRoaW5ncyBzbyB0aGF0IHlvdSBjYW4gcnVuIHRoZSBqb2JzDQo+IGluIHBhcmFs bGVsIGlmIHlvdXIgbGFiIGhhcyBnb3QgYSBjbHVzdGVyIHRvIGRvIHNvLg0KPiBJZg0KPiB0aGlz IGlzIG5vdCB0aGUgY2FzZSwgeWVzLCBpdCBjYW4gYmUgbG9uZywgYW5kIG11Y2gNCj4gbG9uZ2Vy IHRoYW4gZnNsdmJtMiBmb3Igd2hpY2ggeW91IGhhdmUgaG9wZWZ1bGx5DQo+IGNob3NlbiB0aGUg cmVjb21tZW5kZWQgYWZmaW5lIG9wdGlvbiB3aXRoIC1hLg0KPiBKdXN0IGNoZWNrIHRoYXQgeW91 ciA0RCBpbWFnZSBHTV9tb2RfbWVyZy5uaWkuZ3ogaW4NCj4geW91ciBzdGF0cyBkaXJlY3Rvcnkg aGFzIGJlZW4gY3JlYXRlZCBieSB0aGUgbW9zdA0KPiByZWNlbnQgZnNsdmJtMyB5b3UgcmFuLCBh bmQgaWYgc28sIGp1c3QgYWJvcnQgdGhlDQo+IGZpcnN0ICBmc2x2Ym0zIGNvbW1hbmQuIA0KPiAN Cj4gSG9wZSB0aGlzIGhlbHBzLA0KPiBHd2VuYWVsbGUNCj4gDQo+IA0KPiAtLS0gIlpoYW5nLCBY aWFvY2h1IChOSUgvTklEQSkgW0ZdIg0KPiA8emhhbmd4MkBJTlRSQS5OSURBLk5JSC5HT1Y+IGEg w4PCqWNyaXQgOg0KPiANCj4gPiBIaSwNCj4gPiANCj4gPiBJJ20gd29ya2luZyBvbiB0aGUgVkJN IGRhdGEuIE15IHNhbXBsZSBpbmNsdWRlZCA0OA0KPiA+IHN1YmplY3RzICgyNCBwYXRpZW50cw0K PiA+IGFuZCAyNCBjb250cm9scykuIEkgaGF2ZSB0d28gcXVlc3Rpb25zIGFib3V0IHRoZQ0KPiA+ ICJmc2x2Ym1fM19wcm9jIi4NCj4gPiANCj4gPiBUaGUgZmlyc3Qgb25lLCBJIHJhbiBpdCAyIGRh eXMgYWdvLiBIb3dldmVyLCB1cCB0bw0KPiA+IG5vdywgaXQgaGF2ZSBOT1QNCj4gPiBmaW5pc2hl ZC4gSSBjaGVja2VkIHRoZSBwcm9ncmFtIGlzIG5vdCBkaWVkLg0KPiA+IEhvd2V2ZXIsIHdoZW4g SSBydW4NCj4gPiAiZnNsdmJtXzEiIGFuZCAiZnNsdmJtXzIiLCBpdCB0b29rIG5vIG1vcmUgdGhh biA0DQo+ID4gaG91cnMuIElzIGl0IG5vcm1hbA0KPiA+IHBoZW5vbWVub24/DQo+ID4gDQo+ID4g VGhlIHNlY29uZCBvbmUsIGFib3V0IDEgZGF5IGFnbywgbXkgY29tcHV0ZXIgd2FzDQo+ID4gcmVi b290ZWQgYnkgcG93ZXINCj4gPiBwcm9ibGVtLiBBZnRlciByZWJvb3QsIEkgcmUtcnVuIHRoZSAi ZnNsdmJtXzMiLiBJdA0KPiA+IGxvb2tzIGxpa2UgT0suIE5lZWQgSQ0KPiA+IHRvIGRlbGV0ZWQg YWxsIGZpbGVzIGFuZCByZXN0YXJ0IGZyb20gImZzbHZibV8xIj8gDQo+IA0KPiA+IA0KPiA+IFRo YW5rIHlvdSB2ZXJ5IG11Y2ghDQo+ID4gDQo+ID4gWGlhb2NodSBaaGFuZyBQaC5EDQo+ID4gDQo+ ID4gVmlzaXRpbmcgUmVzZWFyY2ggRmVsbG93DQo+ID4gDQo+ID4gTklIL05JREEtSVJQDQo+ID4g DQo+ID4gNTUwMCBOYXRoYW4gU2hvY2sgRHJpdmUNCj4gPiANCj4gPiBCYWx0aW1vcmUgTUQgMjEy MjQNCj4gPiANCj4gPiAgDQo+ID4gDQo+ID4gVGVsOiAoNDEwKSAtIDU1MCAtIDE0NDAgZXh0LiA0 MzQNCj4gPiANCj4gPiAgDQo+ID4gDQo+ID4gDQo+IA0KPiANCj4gDQo+ICAgICAgDQo+DQpfX19f X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19f X19fX19fX19fX19fX19fXw0KPiANCj4gTmUgZ2FyZGV6IHBsdXMgcXUndW5lIHNldWxlIGFkcmVz c2UgbWFpbCAhIENvcGlleg0KPiB2b3MgbWFpbHMgdmVycyBZYWhvbyEgTWFpbCANCj4gDQoNCg0K DQogICAgICBfX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19f X19fX19fX19fX19fX19fX19fX19fX19fX19fXyANCk5lIGdhcmRleiBwbHVzIHF1J3VuZSBzZXVs ZSBhZHJlc3NlIG1haWwgISBDb3BpZXogdm9zIG1haWxzIHZlcnMgWWFob28hIE1haWwgDQo= ========================================================================= Date: Thu, 11 Oct 2007 12:23:28 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Buyean Lee <[log in to unmask]> Subject: Re: FLIRT: Preserver what? In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="--------MB_8C9DA3A311CC182_C3C_A644_FWM-M08.sysops.aol.com" ----------MB_8C9DA3A311CC182_C3C_A644_FWM-M08.sysops.aol.com Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii" Hi Mark, It is very clear to me. Thanks a million. Buyean -----Original Message----- From: Mark Jenkinson <[log in to unmask]> To: [log in to unmask] Sent: Thu, 11 Oct 2007 3:15 am Subject: Re: [FSL] FLIRT: Preserver what? Hi,? ? As I understand it, "concentration" is about local intensity so that conserving? concentration when you are scaling an image (making the structure bigger? or smaller) will keep the intensity values within that region the same.? By comparison, conserving "total (amount)" tries to keep the product of? intensity and volume the same, so that if the volume increases then the? intensity within will decrease.? ? The behaviour of FLIRT is always to keep the intensity the same, thus? this would be "conserving concentration". So you should be fine to use? either flirt, or scripts that call flirt, such as first_flirt.? ? Note that for FMRI analyses, it is common to do grand-mean intensity? normalisation which then rescales the intensities to have a fixed? overall mean across all dimensions (space and time). If such a scaling? is done then it doesn't matter if you tried to correct for the original? flirt scaling or not, as it would be constant across space and time and? therefore would be factored out by this normalisation. Only if you? are wanting to use the intensities in something like VBM would it make? a difference whether you tried to correct for the scaling (preserving? "total").? ? Hope this helps.? All the best,? ? Mark? ? ? On 11 Oct 2007, at 03:35, Buyean Lee wrote:? ? > Hi Steve,? >? > Do you mean FSL normalization will not preserve the concentration?? >? > I am specifically paying attention to this issue due to the > following reason.? >? > I used to normalize the binding potential images of my PET images > using SPM2(5) in which it preserve concentration.? > Since I am not satisfied with the location of the subcortical > regions after normalization, I am looking for a method which can > improve the spatial normalization of the subcortical regions (if > you know one, let me know).? >? > I thought the spatial normalization used in FIRST (i.e., > first_flirt) is promising.? > It seems to work better. But, I don't know if 'first_flirt' will > preserver concentration or not.? >? > Thank you,? >? > Buyean? >? >? > -----Original Message-----? > From: Steve Smith <[log in to unmask]>? > To: [log in to unmask] > Sent: Mon, 8 Oct 2007 11:04 pm? > Subject: Re: [FSL] FLIRT: Preserver what?? >? > Hi,? >? > FLIRT applies an affine-only transformation. If you wish to divide > the output image by the amount of volumetric scaling (to preserve > "concentration") you can extract the volumetric scaling factor > using avscale and then apply this easily with the -mul option > inside fslmaths.? >? > Cheers, Steve.? >? > On 8 Oct 2007, at 22:59, Buyean Lee wrote:? >? > > Dear FSL users,? > >? > > When one normalizes a image in SPM2(5), one can choose which > > parameter (e.g., concentration or total (=amount)) is going to be > > preserved.? > >? > > I read the FSL website and checked FLIRT options, but I can't > find > a similar option.? > >? > > Does FLIRT preserve the concentration?? > >? > > Thank you,? > >? > > Buyean? > > Check Out the new free AIM(R) Mail -- Unlimited storage and > > industry-leading spam and email virus protection.? >? > ----------------------------------------------------------------------> -----? > Stephen M. Smith, Professor of Biomedical Engineering? > Associate Director, Oxford University FMRIB Centre? >? > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK? > +44 (0) 1865 222726 (fax 222717)? > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve? > ----------------------------------------------------------------------> -----? > Check Out the new free AIM(R) Mail -- Unlimited storage and > industry-leading spam and email virus protection.? ________________________________________________________________________ Check Out the new free AIM(R) Mail -- Unlimited storage and industry-leading spam and email virus protection. ----------MB_8C9DA3A311CC182_C3C_A644_FWM-M08.sysops.aol.com Content-Transfer-Encoding: 7bit Content-Type: text/html; charset="us-ascii"
Hi Mark,

It is very clear to me.

Thanks a million.

Buyean



-----Original Message-----
From: Mark Jenkinson <[log in to unmask]>
To: [log in to unmask]
Sent: Thu, 11 Oct 2007 3:15 am
Subject: Re: [FSL] FLIRT: Preserver what?

Hi, 
 
As I understand it, "concentration" is about local intensity so that conserving 
concentration when you are scaling an image (making the structure bigger 
or smaller) will keep the intensity values within that region the same. 
By comparison, conserving "total (amount)" tries to keep the product of 
intensity and volume the same, so that if the volume increases then the 
intensity within will decrease. 
 
The behaviour of FLIRT is always to keep the intensity the same, thus 
this would be "conserving concentration". So you should be fine to use 
either flirt, or scripts that call flirt, such as first_flirt. 
 
Note that for FMRI analyses, it is common to do grand-mean intensity 
normalisation which then rescales the intensities to have a fixed 
overall mean across all dimensions (space and time). If such a scaling 
is done then it doesn't matter if you tried to correct for the original 
flirt scaling or not, as it would be constant across space and time and 
therefore would be factored out by this normalisation. Only if you 
are wanting to use the intensities in something like VBM would it make 
a difference whether you tried to correct for the scaling (preserving 
"total"). 
 
Hope this helps. 
All the best, 
  Mark 
 
 
On 11 Oct 2007, at 03:35, Buyean Lee wrote: 
 
> Hi Steve, 

> Do you mean FSL normalization will not preserve the concentration? 

> I am specifically paying attention to this issue due to the > following reason. 

> I used to normalize the binding potential images of my PET images > using SPM2(5) in which it preserve concentration. 
> Since I am not satisfied with the location of the subcortical > regions after normalization, I am looking for a method which can > improve the spatial normalization of the subcortical regions (if > you know one, let me know). 

> I thought the spatial normalization used in FIRST (i.e., > first_flirt) is promising. 
> It seems to work better. But, I don't know if 'first_flirt' will > preserver concentration or not. 

> Thank you, 

> Buyean 


> -----Original Message----- 
> From: Steve Smith <[log in to unmask]
> To: [log in to unmask] 
> Sent: Mon, 8 Oct 2007 11:04 pm 
> Subject: Re: [FSL] FLIRT: Preserver what? 

> Hi, 

> FLIRT applies an affine-only transformation. If you wish to divide > the output image by the amount of volumetric scaling (to preserve > "concentration") you can extract the volumetric scaling factor > using avscale and then apply this easily with the -mul option > inside fslmaths. 

> Cheers, Steve. 

> On 8 Oct 2007, at 22:59, Buyean Lee wrote: 

> > Dear FSL users, 
> > 
> > When one normalizes a image in SPM2(5), one can choose which > > parameter (e.g., concentration or total (=amount)) is going to be > > preserved. 
> > 
> > I read the FSL website and checked FLIRT options, but I can't > find > a similar option. 
> > 
> > Does FLIRT preserve the concentration? 
> > 
> > Thank you, 
> > 
> > Buyean 
> > Check Out the new free AIM(R) Mail -- Unlimited storage and > > industry-leading spam and email virus protection. 

> ----------------------------------------------------------------------> ----- 
> Stephen M. Smith, Professor of Biomedical Engineering 
> Associate Director, Oxford University FMRIB Centre 

> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK 
> +44 (0) 1865 222726 (fax 222717) 
> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve 
> ----------------------------------------------------------------------> ----- 
> Check Out the new free AIM(R) Mail -- Unlimited storage and > industry-leading spam and email virus protection. 

Check Out the new free AIM(R) Mail -- Unlimited storage and industry-leading spam and email virus protection.
----------MB_8C9DA3A311CC182_C3C_A644_FWM-M08.sysops.aol.com-- ========================================================================= Date: Thu, 11 Oct 2007 18:29:56 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: =?iso-8859-1?q?Gwena=EBlle=20DOUAUD?= <[log in to unmask]> Subject: RE : Re: [FSL] RE : Re: [FSL] RE : [FSL] hi and for VBM problem. In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable If your lab has indeed a cluster (see your IT guys), you could always restart it, it would only overwrite what has been already created. I guess you can stop it by a simple ctrl c in the shell where you've typed the fslvbm3 command. But I would suggest you wait a bit, as it should not be too long to give you the results now.=20 Gwenaelle --- "Zhang, Xiaochu (NIH/NIDA) [F]" <[log in to unmask]> a =E9crit : > Thank you very much!=20 > So my question is whether I am able to interrupt the > program when it is running (For it took a long time, > sometimes it can not be avoid. ) and restart it > again. > Does it influence the final result? > Thanks, again! >=20 > -----Original Message----- > From: Gwena=C3=ABlle DOUAUD > [mailto:[log in to unmask]]=20 > Sent: 2007=E5=B9=B410=E6=9C=8811=E6=97=A5 10:36 > To: [log in to unmask] > Subject: [FSL] RE : Re: [FSL] RE : [FSL] hi and for > VBM problem. >=20 > Hi, >=20 > sorry, I should have been more specific... It can be > run in parallel if your lab has got a SGE system or > something like that (sorry I'm not very good at > these > things) and then you send your job to a bunch of > computers involved in the cluster. It should not > take > much longer anyway.=20 > Hopefully, we should get a better and much quicker > non-linear registration (FNIRT) soonish and that > will > be included in the next FSL4.1 patch. >=20 > Cheers, > Gwenaelle >=20 > =20 > --- "Zhang, Xiaochu (NIH/NIDA) [F]" > <[log in to unmask]> a =C3=A9crit : >=20 > > Thank you very much for speedy response! > > I think the first question has been resolved. For > 2 > > hours per subject, it will take about 100 hours > > (about 4 days) to finish the whole 48 subjects. I > > need wait two days more. > > For you said it can be run in parallel and the > > computer in our lab have multiple CPUs. So, can I > > open multiple terminals and run the script in each > > one to shorten the whole time? > > Thanks again! > >=20 > > -----Original Message----- > > From: Gwena=C3=83=C2=ABlle DOUAUD > > [mailto:[log in to unmask]]=20 > > Sent: 2007=C3=A5=C2=B9=C2=B410=C3=A6=C5=93=CB=8611=C3=A6=E2=80=94=C2=A5= 9:39 > > To: [log in to unmask] > > Subject: [FSL] RE : [FSL] hi and for VBM problem. > >=20 > > Hi, > >=20 > > I'm not sure about your first question. The > > non-linear > > registration can take a while, let's say at least > 2 > > hours per subject.=20 > > We have done the things so that you can run the > jobs > > in parallel if your lab has got a cluster to do > so. > > If > > this is not the case, yes, it can be long, and > much > > longer than fslvbm2 for which you have hopefully > > chosen the recommended affine option with -a. > > Just check that your 4D image GM_mod_merg.nii.gz > in > > your stats directory has been created by the most > > recent fslvbm3 you ran, and if so, just abort the > > first fslvbm3 command.=20 > >=20 > > Hope this helps, > > Gwenaelle > >=20 > >=20 > > --- "Zhang, Xiaochu (NIH/NIDA) [F]" > > <[log in to unmask]> a =C3=83=C2=A9crit : > >=20 > > > Hi, > > >=20 > > > I'm working on the VBM data. My sample included > 48 > > > subjects (24 patients > > > and 24 controls). I have two questions about the > > > "fslvbm_3_proc". > > >=20 > > > The first one, I ran it 2 days ago. However, up > to > > > now, it have NOT > > > finished. I checked the program is not died. > > > However, when I run > > > "fslvbm_1" and "fslvbm_2", it took no more than > 4 > > > hours. Is it normal > > > phenomenon? > > >=20 > > > The second one, about 1 day ago, my computer was > > > rebooted by power > > > problem. After reboot, I re-run the "fslvbm_3". > It > > > looks like OK. Need I > > > to deleted all files and restart from > "fslvbm_1"?=20 > >=20 > > >=20 > > > Thank you very much! > > >=20 > > > Xiaochu Zhang Ph.D > > >=20 > > > Visiting Research Fellow > > >=20 > > > NIH/NIDA-IRP > > >=20 > > > 5500 Nathan Shock Drive > > >=20 > > > Baltimore MD 21224 > > >=20 > > > =20 > > >=20 > > > Tel: (410) - 550 - 1440 ext. 434 > > >=20 > > > =20 > > >=20 > > >=20 > >=20 > >=20 > >=20 > > =20 > > > _________________________________________________________________________= ____ > >=20 > > Ne gardez plus qu'une seule adresse mail ! Copiez > > vos mails vers Yahoo! Mail=20 > >=20 >=20 >=20 >=20 > =20 > _________________________________________________________________________= ____ >=20 > Ne gardez plus qu'une seule adresse mail ! Copiez > vos mails vers Yahoo! Mail=20 >=20 ___________________________________________________________________= __________=20 Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! M= ail=20 ========================================================================= Date: Thu, 11 Oct 2007 13:18:07 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matt Glasser <[log in to unmask]> Organization: ma-tea.com Subject: Re: TBSS-voxelwise correlations between FA and Test scores In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0000_01C80C09.29FAA430" This is a multi-part message in MIME format. ------=_NextPart_000_0000_01C80C09.29FAA430 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Tom and others, I am helping Bhargav with this project. We tried using the -D option, however now all of our t statistics are very small (between -1 and 1, with most very close to zero). This doesn't make sense, however, as we would expect at least some voxels to show t statistics greater than 1, just by chance. Our randomize commandline, our design matrix and our contrast file follow this message. We have verified that our 4D FA skeleton file and mean skeleton mask files are okay. Does anyone know why we are getting such low t values? Would it be possible to get r values instead using something other than randomise? Thanks, Matt. randomise commandline: randomise -i all_FA_skeletonised -o tbss -m mean_FA_skeleton_mask -d stats.mat -t stats.con -n 5000 -c 3 -D stats.mat: /NumWaves 1 /NumPoints 10 /PPheights 7.800000e+02 /Matrix 6.700000e+02 5.400000e+02 5.300000e+02 6.000000e+02 5.800000e+02 7.800000e+02 7.500000e+02 6.100000e+02 7.000000e+02 7.200000e+02 stats.con: /ContrastName1 Positive /ContrastName2 Negative /NumWaves 1 /NumContrasts 2 /PPheights 7.800000e+02 7.800000e+02 /RequiredEffect 3.726 3.726 /Matrix 1.000000e+00 -1.000000e+00 _____ From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Thomas Nichols Sent: Wednesday, October 10, 2007 5:10 AM To: [log in to unmask] Subject: Re: [FSL] TBSS-voxelwise correlations between FA and Test scores Bhargav, Did you include an intercept in the model? Or specify the -D option and use a centered covariate? That might well explain the large t values (without an intercept you may be testing if the FA values are non-zero, not a very interesting question). randomise uses the general linear model and contrasts to test effects of interest with t-tests and so doesn't produce correlation coefficients. You can threshold on the basis of corrected or uncorrected P-values with fslmaths/avwmaths, the -thr option, and the corresponding One-Minus-P-value images; see the randomise help page http://www.fmrib.ox.ac.uk/fsl/randomise/ for the exact filename extensions for each of these types of images. Hope this helps! -Tom On 10/9/07, Bhargav Kumar Errangi <[log in to unmask]> wrote: Hello, We are using TBSS to look for voxel-wise correlations between FA and test scores. Could you please answer the following questions for us? 1) Using the GLM tool, we specificy 1 covariate and provide values for each of 10 subjects in the EV tab. Then in the contrast tab, we specify 2 contrasts (1 and -1) 2) The output is a t statistic image. We were exepcting an r statistic map. Does FSL calculate the t from the r? 3) Will the "1" contrast yield positive correlations and the "-1" contrast yield negative correlations? 4) The t statistic values are very, very large. We only get reasonbable maps when we threshold at t>20. Do you know why this might be? 5) Is there a way to threshold the map based on p-values rather than t statistics? ____________________________________________ Thomas Nichols, PhD Director, Modelling & Genetics GlaxoSmithKline Clinical Imaging Centre Senior Research Fellow Oxford University FMRIB Centre ------=_NextPart_000_0000_01C80C09.29FAA430 Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

Tom and = others,

 

I am helping Bhargav with this project.  We tried using the -D option, however now all of our t statistics are very small (between -1 and 1, with most very close to zero).  This doesn’t make sense, however, as we would expect = at least some voxels to show t statistics greater than 1, just by = chance.  Our randomize commandline, our design matrix and our contrast file follow = this message.  We have verified that our 4D FA skeleton file and mean = skeleton mask files are okay.  Does anyone know why we are getting such low = t values?  Would it be possible to get r values instead using = something other than randomise?

 

Thanks,

=

 

Matt.

 

randomise = commandline:

randomise -i all_FA_skeletonised -o = tbss -m mean_FA_skeleton_mask -d stats.mat -t stats.con -n 5000 -c 3 = -D

 

stats.mat:<= /p>

 

/NumWaves    &nb= sp;  1

/NumPoints    &n= bsp;  10

/PPheights    &n= bsp;           &nb= sp;   7.800000e+02

 

/Matrix

=

6.700000e+02    =

5.400000e+02    =

5.300000e+02    =

6.000000e+02    =

5.800000e+02    =

7.800000e+02    =

7.500000e+02    =

6.100000e+02    =

7.000000e+02    =

7.200000e+02    =

 

stats.con:<= /p>

 

/ContrastName1 = Positive

/ContrastName2 = Negative

/NumWaves    &nb= sp;  1

/NumContrasts  = 2

/PPheights    &n= bsp;           &nb= sp;   7.800000e+02    = 7.800000e+02

/RequiredEffect   &nb= sp;          = 3.726    3.726

 

/Matrix

=

1.000000e+00 =

-1.000000e+00

 


From: = FSL - FMRIB's Software Library [mailto:[log in to unmask]] On = Behalf Of Thomas Nichols
Sent: Wednesday, October = 10, 2007 5:10 AM
To: = [log in to unmask]
Subject: Re: [FSL] = TBSS-voxelwise correlations between FA and Test scores

 

Bhargav,

Did you include an intercept in the model?  Or specify the -D = option and use a centered covariate?  That might well explain the large t = values (without an intercept you may be testing if the FA values are non-zero, = not a very interesting question).

randomise uses the general linear model and contrasts to test effects of interest with t-tests and so doesn't produce correlation = coefficients.

You can threshold on the basis of corrected or uncorrected P-values with fslmaths/avwmaths, the -thr option, and the corresponding = One-Minus-P-value images;  see the randomise help page http://www.fmrib.ox.ac.= uk/fsl/randomise/ for the exact filename extensions for each of these types of images.

Hope this helps!

-Tom

On 10/9/07, Bhargav Kumar Errangi <[log in to unmask]> wrote:

Hello,

We are using TBSS to look for voxel-wise correlations between FA and = test
scores. Could you please answer the following questions for us?

1) Using the GLM tool, we specificy 1 covariate and provide values for = each
of 10 subjects in the EV tab. Then in the contrast tab, we specify 2
contrasts (1 and -1)
2) The output is a t statistic image. We were exepcting an r statistic = map.
Does FSL calculate the t from the r?
3) Will the "1" contrast yield positive correlations and the "-1" contrast
yield negative correlations?
4) The t statistic values are very, very large. We only get reasonbable = maps
when we threshold at t>20. Do you know why this might be?
5) Is there a way to threshold the map based on p-values rather than t =
statistics?



____________________________________________
Thomas Nichols, PhD
Director, Modelling & Genetics
GlaxoSmithKline Clinical Imaging Centre

Senior = Research Fellow
Oxford University
FMRIB Centre

------=_NextPart_000_0000_01C80C09.29FAA430-- ========================================================================= Date: Thu, 11 Oct 2007 19:31:35 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: =?iso-8859-1?q?Gwena=EBlle=20DOUAUD?= <[log in to unmask]> Subject: RE : Re: [FSL] TBSS-voxelwise correlations between FA and Test scores In-Reply-To: [log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Hi Matt, it seems that you have not demeaned your ev in your stats.mat. You should do so and run exactly the same command you've used without forgetting the -D option. Cheers, Gwenaelle --- Matt Glasser <[log in to unmask]> a =E9crit : > Tom and others, >=20 > =20 >=20 > I am helping Bhargav with this project. We tried > using the -D option, > however now all of our t statistics are very small > (between -1 and 1, with > most very close to zero). This doesn't make sense, > however, as we would > expect at least some voxels to show t statistics > greater than 1, just by > chance. Our randomize commandline, our design > matrix and our contrast file > follow this message. We have verified that our 4D > FA skeleton file and mean > skeleton mask files are okay. Does anyone know why > we are getting such low > t values? Would it be possible to get r values > instead using something > other than randomise? >=20 > =20 >=20 > Thanks, >=20 > =20 >=20 > Matt. >=20 > =20 >=20 > randomise commandline: >=20 > randomise -i all_FA_skeletonised -o tbss -m > mean_FA_skeleton_mask -d > stats.mat -t stats.con -n 5000 -c 3 -D >=20 > =20 >=20 > stats.mat: >=20 > =20 >=20 > /NumWaves 1 >=20 > /NumPoints 10 >=20 > /PPheights 7.800000e+02 >=20 > =20 >=20 > /Matrix >=20 > 6.700000e+02 =20 >=20 > 5.400000e+02 =20 >=20 > 5.300000e+02 =20 >=20 > 6.000000e+02 =20 >=20 > 5.800000e+02 =20 >=20 > 7.800000e+02 =20 >=20 > 7.500000e+02 =20 >=20 > 6.100000e+02 =20 >=20 > 7.000000e+02 =20 >=20 > 7.200000e+02 =20 >=20 > =20 >=20 > stats.con: >=20 > =20 >=20 > /ContrastName1 Positive >=20 > /ContrastName2 Negative >=20 > /NumWaves 1 >=20 > /NumContrasts 2 >=20 > /PPheights 7.800000e+02 =20 > 7.800000e+02 >=20 > /RequiredEffect 3.726 3.726 >=20 > =20 >=20 > /Matrix >=20 > 1.000000e+00=20 >=20 > -1.000000e+00 >=20 > =20 >=20 > _____ =20 >=20 > From: FSL - FMRIB's Software Library > [mailto:[log in to unmask]] On Behalf > Of Thomas Nichols > Sent: Wednesday, October 10, 2007 5:10 AM > To: [log in to unmask] > Subject: Re: [FSL] TBSS-voxelwise correlations > between FA and Test scores >=20 > =20 >=20 > Bhargav, >=20 > Did you include an intercept in the model? Or > specify the -D option and use > a centered covariate? That might well explain the > large t values (without > an intercept you may be testing if the FA values are > non-zero, not a very > interesting question). >=20 > randomise uses the general linear model and > contrasts to test effects of > interest with t-tests and so doesn't produce > correlation coefficients. >=20 > You can threshold on the basis of corrected or > uncorrected P-values with > fslmaths/avwmaths, the -thr option, and the > corresponding One-Minus-P-value > images; see the randomise help page > http://www.fmrib.ox.ac.uk/fsl/randomise/ for the > exact filename extensions > for each of these types of images. >=20 > Hope this helps! >=20 > -Tom >=20 > On 10/9/07, Bhargav Kumar Errangi > <[log in to unmask]> wrote: >=20 > Hello, >=20 > We are using TBSS to look for voxel-wise > correlations between FA and test > scores. Could you please answer the following > questions for us? >=20 > 1) Using the GLM tool, we specificy 1 covariate and > provide values for each=20 > of 10 subjects in the EV tab. Then in the contrast > tab, we specify 2 > contrasts (1 and -1) > 2) The output is a t statistic image. We were > exepcting an r statistic map. > Does FSL calculate the t from the r? > 3) Will the "1" contrast yield positive correlations > and the "-1" contrast=20 > yield negative correlations? > 4) The t statistic values are very, very large. We > only get reasonbable maps > when we threshold at t>20. Do you know why this > might be? > 5) Is there a way to threshold the map based on > p-values rather than t=20 > statistics? >=20 >=20 >=20 > ____________________________________________ > Thomas Nichols, PhD > Director, Modelling & Genetics > GlaxoSmithKline Clinical Imaging Centre >=20 > Senior Research Fellow=20 > Oxford University FMRIB Centre=20 >=20 >=20 ___________________________________________________________________= __________=20 Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! M= ail=20 ========================================================================= Date: Thu, 11 Oct 2007 19:53:39 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Bhargav Kumar Errangi <[log in to unmask]> Subject: Re: RE : Re: [FSL] TBSS-voxelwise correlations between FA and Test scores Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Thank you for your response. I am not sure that I know what is meant by "demean". We did use the -D option, but it seems you are suggesting we do something additional with t= he covariate in the design matrix. Can you please describe what we should do= ? Thank you very much, Bhargav Gwena=EBlle DOUAUD <[log in to unmask]> wrote: Hi Matt, it seems that you have not demeaned your ev in your stats.mat. You should do so and run exactly the same command you've used without forgetting the -D option. Cheers, Gwenaelle --- Matt Glasser a =E9crit : > Tom and others, > > > > I am helping Bhargav with this project. We tried > using the -D option, > however now all of our t statistics are very small > (between -1 and 1, with > most very close to zero). This doesn't make sense, > however, as we would > expect at least some voxels to show t statistics > greater than 1, just by > chance. Our randomize commandline, our design > matrix and our contrast file > follow this message. We have verified that our 4D > FA skeleton file and mean > skeleton mask files are okay. Does anyone know why > we are getting such low > t values? Would it be possible to get r values > instead using something > other than randomise? > > > > Thanks, > > > > Matt. > > > > randomise commandline: > > randomise -i all_FA_skeletonised -o tbss -m > mean_FA_skeleton_mask -d > stats.mat -t stats.con -n 5000 -c 3 -D > > > > stats.mat: > > > > /NumWaves 1 > > /NumPoints 10 > > /PPheights 7.800000e+02 > > > > /Matrix > > 6.700000e+02 > > 5.400000e+02 > > 5.300000e+02 > > 6.000000e+02 > > 5.800000e+02 > > 7.800000e+02 > > 7.500000e+02 > > 6.100000e+02 > > 7.000000e+02 > > 7.200000e+02 > > > > stats.con: > > > > /ContrastName1 Positive > > /ContrastName2 Negative > > /NumWaves 1 > > /NumContrasts 2 > > /PPheights 7.800000e+02 > 7.800000e+02 > > /RequiredEffect 3.726 3.726 > > > > /Matrix > > 1.000000e+00 > > -1.000000e+00 > > > > _____ > > From: FSL - FMRIB's Software Library > [mailto:[log in to unmask]] On Behalf > Of Thomas Nichols > Sent: Wednesday, October 10, 2007 5:10 AM > To: [log in to unmask] > Subject: Re: [FSL] TBSS-voxelwise correlations > between FA and Test scores > > > > Bhargav, > > Did you include an intercept in the model? Or > specify the -D option and use > a centered covariate? That might well explain the > large t values (without > an intercept you may be testing if the FA values are > non-zero, not a very > interesting question). > > randomise uses the general linear model and > contrasts to test effects of > interest with t-tests and so doesn't produce > correlation coefficients. > > You can threshold on the basis of corrected or > uncorrected P-values with > fslmaths/avwmaths, the -thr option, and the > corresponding One-Minus-P-value > images; see the randomise help page > http://www.fmrib.ox.ac.uk/fsl/randomise/ for the > exact filename extensions > for each of these types of images. > > Hope this helps! > > -Tom > > On 10/9/07, Bhargav Kumar Errangi > wrote: > > Hello, > > We are using TBSS to look for voxel-wise > correlations between FA and test > scores. Could you please answer the following > questions for us? > > 1) Using the GLM tool, we specificy 1 covariate and > provide values for each > of 10 subjects in the EV tab. Then in the contrast > tab, we specify 2 > contrasts (1 and -1) > 2) The output is a t statistic image. We were > exepcting an r statistic map. > Does FSL calculate the t from the r? > 3) Will the "1" contrast yield positive correlations > and the "-1" contrast > yield negative correlations? > 4) The t statistic values are very, very large. We > only get reasonbable maps > when we threshold at t>20. Do you know why this > might be? > 5) Is there a way to threshold the map based on > p-values rather than t > statistics? > > > > ____________________________________________ > Thomas Nichols, PhD > Director, Modelling & Genetics > GlaxoSmithKline Clinical Imaging Centre > > Senior Research Fellow > Oxford University FMRIB Centre ========================================================================= Date: Thu, 11 Oct 2007 21:07:07 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: =?iso-8859-1?q?Gwena=EBlle=20DOUAUD?= <[log in to unmask]> Subject: RE : Re: [FSL] RE : Re: [FSL] TBSS-voxelwise correlations between FA and Test scores In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Sure. By that I mean that you have to centre your ev around its mean. You need to demean (remove the mean) both your data - which is done by the -D option - AND your covariate. Assuming that the vector y is your covariate, then you just need to do y-mean(y), which gives you the demeaned covariate: 22 -108 -118 -48 -68 132 102 -38 52 72 Cheers, Gwen --- Bhargav Kumar Errangi <[log in to unmask]> a =E9crit : > Thank you for your response. >=20 > I am not sure that I know what is meant by "demean". > We did use the -D > option, but it seems you are suggesting we do > something additional with the > covariate in the design matrix. Can you please > describe what we should do? >=20 > Thank you very much, >=20 > Bhargav >=20 > Gwena=EBlle DOUAUD <[log in to unmask]> wrote: >=20 > Hi Matt, >=20 > it seems that you have not demeaned your ev in > your > stats.mat. You should do so and run exactly the > same > command you've used without forgetting the -D > option. >=20 > Cheers, > Gwenaelle >=20 > --- Matt Glasser a =E9crit : >=20 > > Tom and others, > > > > > > > > I am helping Bhargav with this project. We > tried > > using the -D option, > > however now all of our t statistics are very > small > > (between -1 and 1, with > > most very close to zero). This doesn't make > sense, > > however, as we would > > expect at least some voxels to show t > statistics > > greater than 1, just by > > chance. Our randomize commandline, our design > > matrix and our contrast file > > follow this message. We have verified that our > 4D > > FA skeleton file and mean > > skeleton mask files are okay. Does anyone know > why > > we are getting such low > > t values? Would it be possible to get r values > > instead using something > > other than randomise? > > > > > > > > Thanks, > > > > > > > > Matt. > > > > > > > > randomise commandline: > > > > randomise -i all_FA_skeletonised -o tbss -m > > mean_FA_skeleton_mask -d > > stats.mat -t stats.con -n 5000 -c 3 -D > > > > > > > > stats.mat: > > > > > > > > /NumWaves 1 > > > > /NumPoints 10 > > > > /PPheights 7.800000e+02 > > > > > > > > /Matrix > > > > 6.700000e+02 > > > > 5.400000e+02 > > > > 5.300000e+02 > > > > 6.000000e+02 > > > > 5.800000e+02 > > > > 7.800000e+02 > > > > 7.500000e+02 > > > > 6.100000e+02 > > > > 7.000000e+02 > > > > 7.200000e+02 > > > > > > > > stats.con: > > > > > > > > /ContrastName1 Positive > > > > /ContrastName2 Negative > > > > /NumWaves 1 > > > > /NumContrasts 2 > > > > /PPheights 7.800000e+02 > > 7.800000e+02 > > > > /RequiredEffect 3.726 3.726 > > > > > > > > /Matrix > > > > 1.000000e+00 > > > > -1.000000e+00 > > > > > > > > _____ > > > > From: FSL - FMRIB's Software Library > > [mailto:[log in to unmask]] On Behalf > > Of Thomas Nichols > > Sent: Wednesday, October 10, 2007 5:10 AM > > To: [log in to unmask] > > Subject: Re: [FSL] TBSS-voxelwise correlations > > between FA and Test scores > > > > > > > > Bhargav, > > > > Did you include an intercept in the model? Or > > specify the -D option and use > > a centered covariate? That might well explain > the > > large t values (without > > an intercept you may be testing if the FA > values are > > non-zero, not a very > > interesting question). > > > > randomise uses the general linear model and > > contrasts to test effects of > > interest with t-tests and so doesn't produce > > correlation coefficients. > > > > You can threshold on the basis of corrected or > > uncorrected P-values with > > fslmaths/avwmaths, the -thr option, and the > > corresponding One-Minus-P-value > > images; see the randomise help page > > http://www.fmrib.ox.ac.uk/fsl/randomise/ for > the > > exact filename extensions > > for each of these types of images. > > > > Hope this helps! > > > > -Tom > > > > On 10/9/07, Bhargav Kumar Errangi > > wrote: > > > > Hello, > > > > We are using TBSS to look for voxel-wise > > correlations between FA and test > > scores. Could you please answer the following > > questions for us? > > > > 1) Using the GLM tool, we specificy 1 > covariate and > > provide values for each > > of 10 subjects in the EV tab. Then in the > contrast > > tab, we specify 2 > > contrasts (1 and -1) > > 2) The output is a t statistic image. We were > > exepcting an r statistic map. > > Does FSL calculate the t from the r? > > 3) Will the "1" contrast yield positive > correlations > > and the "-1" contrast > > yield negative correlations? > > 4) The t statistic values are very, very > large. We > > only get reasonbable maps > > when we threshold at t>20. Do you know why > this > > might be? > > 5) Is there a way to threshold the map based > on > > p-values rather than t > > statistics? > > > > > > > > ____________________________________________ > > Thomas Nichols, PhD > > Director, Modelling & Genetics > > GlaxoSmithKline Clinical Imaging Centre > > > > Senior Research Fellow > > Oxford University FMRIB Centre >=20 ___________________________________________________________________= __________=20 Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! M= ail=20 ========================================================================= Date: Thu, 11 Oct 2007 21:38:39 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: =?ISO-8859-2?Q?Ji=F8=ED_Keller?= <[log in to unmask]> Subject: dtifit throwing instance MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hello, I have just got to my first diffusion data analysis, after (hopefully correct ;-) ) creation of bvec and bvals, I ran eddy_correct, extracted B0 image using fslsplit (OK, it CAN be done better ;) ), beted the image and I was looking forward to dtifit output. After running: dtifit -k br-corrected.nii.gz -o dti -m nodif_brain_mask -r bvecs -b bvals -V following output appeared: ---------------------------------- data file br-corrected.nii.gz mask file nodif_brain_mask bvecs bvecs bvals bvals reading data reading mask ok 0 192 0 192 0 70 setting up vols copying input properties to output volumes zeroing output volumes ok Forming A matrix starting the fits 0 slices processed 1 slices processed 2 slices processed 3 slices processed 4 slices processed 5 slices processed 6 slices processed 7 slices processed 8 slices processed 9 slices processed 10 slices processed terminate called after throwing an instance of 'NEWMAT::IncompatibleDimensionsException' Aborted -------------------- Any idea what went wrong ? With thanks George ========================================================================= Date: Thu, 11 Oct 2007 20:57:25 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Brian Patenaude <[log in to unmask]> Subject: Re: first_utils In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hi, Given A design matrix: EV0 EV1 EV2 EV3 1 1 10 50 1 1 20 90 1 -1 10 200 1 -1 2 5 1 1 4 5 1 -1 30 8 first_utils will demean EV1,EV2,EV3. If the ones column is not in the design matrix first_utils will demean the other EVs and add the ones column. So, to test for an effect on each EV, the contrasts for EV1, EV2, EV3 would be [0 1 0 0], [0 0 1 0], [0 0 0 1] respectively. first_utils by default outputs the results for each contrast (specified by the 1,2,3 appended to the file name). Currently, first_utils does not allow you to specify a contrast. For the surface output, the mean surfaces correspond to EV1 and are the same for each contrast. The group membership from which the surface means are evaluated are based on those less than zero versus those that are greater than zero (for EV1 after demeaning). Hopefully this clear thing up, Cheers, Brian > sorry to keep piling up the questions, but if i set up my regression for > first_utils as one group, with one regressor, then only try to look at > that regressor (just like i would in feat), where does first_utils split > them to create the 2 means displayed in the .vtk? > > ie: > > EV1 EV2 > 1 10 > 1 5 > 1 30 > 1 12 > > contrast: > 0 1 ========================================================================= Date: Fri, 12 Oct 2007 11:14:51 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Tim Behrens <[log in to unmask]> Subject: Re: dtifit throwing instance In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=UTF-8; delsp=yes; format=flowed Content-Transfer-Encoding: quoted-printable can you run bedpostx_datacheck on the directory? Cheers T On 11 Oct 2007, at 20:38, Ji=C5=99=C3=AD Keller wrote: > Hello, > I have just got to my first diffusion data analysis, after > (hopefully correct ;-) ) creation of bvec and bvals, I ran > eddy_correct, extracted B0 image using fslsplit (OK, it CAN be done > better ;) ), beted the image and I was looking forward to dtifit > output. > After running: > > dtifit -k br-corrected.nii.gz -o dti -m nodif_brain_mask -r bvecs -=20= > b bvals -V > > following output appeared: > ---------------------------------- > > data file br-corrected.nii.gz > mask file nodif_brain_mask > bvecs bvecs > bvals bvals > reading data > reading mask > ok > 0 192 0 192 0 70 > setting up vols > copying input properties to output volumes > zeroing output volumes > ok > Forming A matrix > starting the fits > 0 slices processed > 1 slices processed > 2 slices processed > 3 slices processed > 4 slices processed > 5 slices processed > 6 slices processed > 7 slices processed > 8 slices processed > 9 slices processed > 10 slices processed > terminate called after throwing an instance of > 'NEWMAT::IncompatibleDimensionsException' > Aborted > -------------------- > Any idea what went wrong ? > > With thanks > George ========================================================================= Date: Fri, 12 Oct 2007 14:34:27 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Pia Rotshtein <[log in to unmask]> Subject: FW: lecture/senior lecture/redear in Edinburgh(job ad) MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01C80CD4.9B853C3C" ------_=_NextPart_001_01C80CD4.9B853C3C Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable =20 =20 ________________________________ From: Alexa Morcom [mailto:[log in to unmask]]=20 Sent: 12 October 2007 13:30 To: [log in to unmask] Subject: Favour (job ad) =20 =20 Senior Lecturer/Reader Cognitive Neuroscience =A342,791 - =A348,161=20 The University of Edinburgh invites applications for a Senior = Lecturer/Reader (clinical or non-clinical) in the School of Biomedical = Sciences to be based in the Centre for Cognitive and Neural Systems.=20 This post is primarily intended for someone with interests and = experience in human brain imaging who would interact with both members = of the Centre for Cognitive and Neural Systems (Professor Richard = Morris, FRS) and Human Cognitive Neuroscience Group (Professor Robert = Logie). Those with interests and skills in relevant aspects of = behavioural neuroscience, neuroanatomy or imaging neuroscience will also = be considered.=20 Applicants should have a PhD or equivalent experience. Preference will = be given to those candidates who have a record of good publications and = a proven record in applying for, and success in obtaining research = funding. Applications from both non-clinical and clinical candidates are = encouraged, and there is flexibility for an appointment to be made at a = more junior level if appropriate. Informal enquiries should be made to Professor R G M Morris FRS ( = [log in to unmask] ).=20 Interviews will be held in January 2008.=20 See http://www.nature.com/naturejobs/science/jobs/28208 = and = http://www.jobs.ed.ac.uk , ref: 3008065 for = full details Closing date: 1 November 2007=20 =20 ------_=_NextPart_001_01C80CD4.9B853C3C Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

 =

 =


From: Alexa Morcom [mailto:[log in to unmask]]
Sent: 12 October 2007 = 13:30
To: = [log in to unmask]
Subject: Favour (job = ad)

 

 

Senior = Lecturer/Reader

Cognitive = Neuroscience

=A342,791 - = =A348,161 =

The University of Edinburgh invites = applications for a Senior Lecturer/Reader (clinical or non-clinical) in the School of Biomedical Sciences to be based in the Centre for = Cognitive and Neural Systems.

This post is primarily intended = for someone with interests and experience in human brain imaging who would interact = with both members of the Centre for Cognitive and Neural Systems (Professor = Richard Morris, FRS) and Human Cognitive Neuroscience Group (Professor Robert = Logie). Those with interests and skills in relevant aspects of behavioural neuroscience, neuroanatomy or imaging neuroscience will also be = considered.

Applicants should have a PhD or = equivalent experience. Preference will be given to those candidates who have a = record of good publications and a proven record in applying for, and success in = obtaining research funding. Applications from both non-clinical and clinical = candidates are encouraged, and there is flexibility for an appointment to be made = at a more junior level if appropriate.

Informal enquiries should be = made to Professor R G M Morris FRS ( [log in to unmask] ). =

Interviews will be held in = January 2008.

See http://www.= nature.com/naturejobs/science/jobs/28208 and http://www.jobs.ed.ac.uk, ref: 3008065 for full details

Closing date: 1 November 2007 =

 

------_=_NextPart_001_01C80CD4.9B853C3C-- ========================================================================= Date: Fri, 12 Oct 2007 14:47:31 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Duncan Mortimer <[log in to unmask]> Subject: Re: running read_avw commands in matlab In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: multipart/alternative; boundary=Apple-Mail-5--858713187 --Apple-Mail-5--858713187 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed What OS are you using? The problem is that the environment variable FSLDIR isn't being set in the Matlab shell. BASH is the recommended shell, and isn't the problem. Could you issue the following in a fresh BASH shell: echo $FSLDIR If this doesn't print out the location of your FSL installation then this is the problem, and you need to go and check the install documentation for setting up this variable. If it returns the expected directory, launch Matlab and check that the FSLDIR environment variable is still set: fsldir=getenv('FSLDIR') Duncan On 11 Oct 2007, at 16:39, Ted Yanagihara wrote: > Hello list, > > I am trying to read some of my images from FEAT and probtrackx in > matlab. I have been getting an error when using read_avw and > read_avw_complex, but the other commands (such as read_avw_hdr) > work fine. > > After looking at the scripts for these functions, it appears that > matlab cannot find the path for fsl.sh because it wants to use the > script says to use this path: ${FSLDIR}/etc/fslconf/fsl.sh but > matlab reads this as: /etc/fslconf/fsl.sh > > Is this problem because matlab is running from bash as its default > when I need it to run in tcsh? If so, does anyone know how I can > make my default environment tcsh? If not, what else might be going on? > > > Here is the full error that matlab gives me when I run this set of > commands: > > >> dir='~/Lab/Data/AKL_13/dti_FA'; > >> fid=fopen(dir); > >> image=read_avw(fid) > Warning: The argument for the %s format specifier must be of type > char (a string). > > In read_avw at 30 > sh: line 1: /etc/fslconf/fsl.sh: No such file or directory > ??? Error using ==> fread > Invalid file identifier. Use fopen to generate a valid file > identifier. > > Error in ==> read_avw_hdr at 17 > testval = fread(fid,1,'int32'); > > Error in ==> read_avw at 33 > [dims,scales,bpp,endian,datatype]= read_avw_hdr(tmpname); WWW: http://www.fmrib.ox.ac.uk/~duncan email: [log in to unmask] --Apple-Mail-5--858713187 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1
What OS are you = using?
The problem is that the environment variable FSLDIR isn't = being set in the Matlab shell.
BASH is the recommended shell, and = isn't the problem. Could you issue the following in a fresh BASH = shell:

echo = $FSLDIR

If = this doesn't print out the location of your FSL installation then this = is the problem, and you need to go and check the install documentation = for setting up this variable.

If it returns the expected = directory, launch Matlab and check that the FSLDIR environment variable = is still set:

fsldir=3Dgetenv('FSLDIR')

Duncan

On 11 Oct 2007, at 16:39, Ted Yanagihara wrote:

Hello = list,

I am trying to read some of my images from FEAT and = probtrackx in matlab. I have been getting an error when using read_avw = and read_avw_complex, but the other commands (such as read_avw_hdr) work = fine.

After looking at the scripts for these functions, it = appears that matlab cannot find the path for fsl.sh because it wants to = use the script says to use this path: ${FSLDIR}/etc/fslconf/fsl.sh=A0=A0 = but matlab reads this as: /etc/fslconf/fsl.sh

Is this problem = because matlab is running from bash as its default when I need it to run = in tcsh? If so, does anyone know how I can make my default environment = tcsh? If not, what else might be going on?


Here is the full = error that matlab gives me when I run this set of = commands:

>> dir=3D'~/Lab/Data/AKL_13/dti_FA';
>> = fid=3Dfopen(dir);
>> image=3Dread_avw(fid)
Warning: The = argument for the %s format specifier must be of type char (a string). =
> In read_avw at 30
sh: line 1: /etc/fslconf/fsl.sh: No such = file or directory
??? Error using =3D=3D> fread
Invalid file = identifier.=A0 Use fopen to generate a valid file = identifier.

Error in =3D=3D> read_avw_hdr at 17
testval =3D = fread(fid,1,'int32');

Error in =3D=3D> read_avw at 33
=A0 = [dims,scales,bpp,endian,datatype]=3D = read_avw_hdr(tmpname);



= --Apple-Mail-5--858713187-- ========================================================================= Date: Fri, 12 Oct 2007 15:59:45 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Vishwadeep Ahluwalia <[log in to unmask]> Subject: ICA-component Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Hi, (resent as new topic) I have 202 volumes of spiral in/out data with TR=3D2s. i ran melodic on t= his data. According to the previous mail in this topic, the highest freq that= can be represented corresponds to 101 cycles. Ive 8 blocks i.e 8 cycles, = so the 8th value in the text file should have a large value for the component corresponding to my expt., which i do get in many components. However in atleast 2 components( this is true in other datasets as well) = i get a peak at 0.15hz(~60cycles in 404s). Is this the peak corresponding t= o respiration? Also, since my sampling frequency is 0.5Hz(corresponds to 20= 2 cycles in 404s), and the cardiac pulsation being much more than that, wil= l i see the peak at the aliased frequency or will it not be shown at all? Iv= e uploaded the times series ,power spectrum and the ICmap of the 0.15Hz.The= upload id is 749962. Thanks -Vish ========================================================================= Date: Fri, 12 Oct 2007 16:08:14 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mahinda Yogarajah <[log in to unmask]> Subject: Re: probtrackx error Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Thank you Ted, Saad, and Matthew for all your help and suggestions (+patience) - probtrackx is working now. Mahinda ========================================================================= Date: Fri, 12 Oct 2007 11:37:20 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Ted Yanagihara <[log in to unmask]> Subject: Re: running read_avw commands in matlab In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_22824_27151095.1192203440707" ------=_Part_22824_27151095.1192203440707 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi Duncan and the list, Our systems administrator has found a solution, or perhaps a workaround, to the problem. I am using mac OSX that runs bash as the default. He changed this default in Netinfo Manager so that I startup in tcsh, then added the necessary information to keep running fsl commands from my bash profile into the tcsh file. Now when I check the environment that Matlab is running in using the command you suggested, getenv('FSLDIR'), it returns the appropriate "/usr/local/fsl" that I was looking for. I hope this helps anyone else who may run into this problem. I know of at least one other person who has struggled with this on a different system. thanks for your help, ted On 10/12/07, Duncan Mortimer <[log in to unmask]> wrote: > > What OS are you using? > The problem is that the environment variable FSLDIR isn't being set in the > Matlab shell.BASH is the recommended shell, and isn't the problem. Could > you issue the following in a fresh BASH shell: > > echo $FSLDIR > > If this doesn't print out the location of your FSL installation then this > is the problem, and you need to go and check the install documentation for > setting up this variable. > > If it returns the expected directory, launch Matlab and check that the > FSLDIR environment variable is still set: > > fsldir=getenv('FSLDIR') > > Duncan > > On 11 Oct 2007, at 16:39, Ted Yanagihara wrote: > > Hello list, > > I am trying to read some of my images from FEAT and probtrackx in matlab. > I have been getting an error when using read_avw and read_avw_complex, but > the other commands (such as read_avw_hdr) work fine. > > After looking at the scripts for these functions, it appears that matlab > cannot find the path for fsl.sh because it wants to use the script says to > use this path: ${FSLDIR}/etc/fslconf/fsl.sh but matlab reads this as: > /etc/fslconf/fsl.sh > > Is this problem because matlab is running from bash as its default when I > need it to run in tcsh? If so, does anyone know how I can make my default > environment tcsh? If not, what else might be going on? > > > Here is the full error that matlab gives me when I run this set of > commands: > > >> dir='~/Lab/Data/AKL_13/dti_FA'; > >> fid=fopen(dir); > >> image=read_avw(fid) > Warning: The argument for the %s format specifier must be of type char (a > string). > > In read_avw at 30 > sh: line 1: /etc/fslconf/fsl.sh: No such file or directory > ??? Error using ==> fread > Invalid file identifier. Use fopen to generate a valid file identifier. > > Error in ==> read_avw_hdr at 17 > testval = fread(fid,1,'int32'); > > Error in ==> read_avw at 33 > [dims,scales,bpp,endian,datatype]= read_avw_hdr(tmpname); > > > > WWW: http://www.fmrib.ox.ac.uk/~duncan > email: [log in to unmask] > > > > ------=_Part_22824_27151095.1192203440707 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi Duncan and the list,

Our systems administrator has found a solution, or perhaps a workaround, to the problem. I am using mac OSX that runs bash as the default. He changed this default in Netinfo Manager so that I startup in tcsh, then added the necessary information to keep running fsl commands from my bash profile into the tcsh file. Now when I check the environment that Matlab is running in using the command you suggested, getenv('FSLDIR'), it returns the appropriate "/usr/local/fsl" that I was looking for.

I hope this helps anyone else who may run into this problem. I know of at least one other person who has struggled with this on a different system.

thanks for your help,

ted

On 10/12/07, Duncan Mortimer <[log in to unmask]> wrote:
What OS are you using?
The problem is that the environment variable FSLDIR isn't being set in the Matlab shell.
BASH is the recommended shell, and isn't the problem. Could you issue the following in a fresh BASH shell:

echo $FSLDIR

If this doesn't print out the location of your FSL installation then this is the problem, and you need to go and check the install documentation for setting up this variable.

If it returns the expected directory, launch Matlab and check that the FSLDIR environment variable is still set:

fsldir=getenv('FSLDIR')

Duncan

On 11 Oct 2007, at 16:39, Ted Yanagihara wrote:

Hello list,

I am trying to read some of my images from FEAT and probtrackx in matlab. I have been getting an error when using read_avw and read_avw_complex, but the other commands (such as read_avw_hdr) work fine.

After looking at the scripts for these functions, it appears that matlab cannot find the path for fsl.sh because it wants to use the script says to use this path: ${FSLDIR}/etc/fslconf/fsl.sh   but matlab reads this as: /etc/fslconf/fsl.sh

Is this problem because matlab is running from bash as its default when I need it to run in tcsh? If so, does anyone know how I can make my default environment tcsh? If not, what else might be going on?


Here is the full error that matlab gives me when I run this set of commands:

>> dir='~/Lab/Data/AKL_13/dti_FA';
>> fid=fopen(dir);
>> image=read_avw(fid)
Warning: The argument for the %s format specifier must be of type char (a string).
> In read_avw at 30
sh: line 1: /etc/fslconf/fsl.sh: No such file or directory
??? Error using ==> fread
Invalid file identifier.  Use fopen to generate a valid file identifier.

Error in ==> read_avw_hdr at 17
testval = fread(fid,1,'int32');

Error in ==> read_avw at 33
  [dims,scales,bpp,endian,datatype]= read_avw_hdr(tmpname);




------=_Part_22824_27151095.1192203440707-- ========================================================================= Date: Fri, 12 Oct 2007 16:47:06 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Tamara Cristescu <[log in to unmask]> Subject: volume of GM/WM/CSF from ROI Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Hi, I would like to extract the amount of WM, GM and CSF from a hippocampal R= OI. I have a T2 rat brain acquired at 0.160 x 0.160 x 0.320 (TE 55). I multip= ly the voxel size by 6 to get to 0.960 x 0.960 x 1.920 (otherwise I can't do= anything with the data in FSL because voxels are too small).=20 I follow the standard analysis protocol: Bet, estimate bias field + bias field correction, and FAST on corrected volume. After FAST I get rest1 I = get the pve files for each tissue (prest1_pve_1=3DGM, rest1_pve_2=3DWM and rest1_pve_0=3DCSF) and the seg files (rest1_seg_1, rest1_seg_2 and rest1_seg_0) and the the whole brain segmented image rest1_seg. (rest1 is= the name of the volulme). I create a hippocampus mask called hippo-mask (intensity 1) and use fslstats to extract the info for GM: fslstats rest1_pve_1 -l 1 -u 1 -k hippo-mask -m=20 0.913644 fslstats rest1_pve_1 -l 1 -u 1 -k hippo-mask -v 6873 12161.580241 I do the same for WM rest1_pve_2 (l=3D2 and u=3D2) and for CSF rest1_pve_= 0 (l=3D0, u=3D0). I multiply the mean value (0.913644) with the 6873 to get the volume in voxels - 6,279 - or multiply the mean value 6873 with 12161.580241 to get= the volume in cubic mm - 11,103. The values seem a little bit unusual to = me for a small rat hippocampus.=20 I would like to extract WM, GM and CSF from the hipopcampus and whole bra= in volume to put them into a volumetric analysis.=20 I would be greateful for any suggestion on how to deal with this data or = any alternatives for the volumetric measurement.=20 Tamara ========================================================================= Date: Fri, 12 Oct 2007 17:32:04 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Christian F. Beckmann" <[log in to unmask]> Subject: Re: ICA - rescaling x-axis of FFT of TC In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: multipart/alternative; boundary=Apple-Mail-18--848840480 --Apple-Mail-18--848840480 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Hi it's possible that this is respiration, though that's be ~13sec per respiratory cycle. The cardiac effect might show up, but at the aliased frequency, see the ssection on respiration and cardiac effects in http://www.fmrib.ox.ac.uk/analysis/techrep/tr05cb1/ tr05cb1.pdf cheers Christian On 10 Oct 2007, at 19:13, Vishwadeep Ahluwalia wrote: > Hi, > I have 202 volumes of spiral in/out data with TR=2s. i ran melodic > on this > data. According to the previous mail in this topic, the highest > freq that > can be represented corresponds to 101 cycles. Ive 8 blocks i.e 8 > cycles, so > the 8th value in the text file should have a large value for > the > component corresponding to my expt., which i do get in many > components. However in atleast 2 components( this is true in > other > datasets as well) i get a peak at 0.15hz(~60cycles in 404s). > Is this > the peak corresponding to respiration? Also, since my sampling > frequency is 0.5Hz(corresponds to 202 cycles in 404s), and the > cardiac pulsation being much more than that, will i see the > peak at > the aliased frequency or will it not be shown at all? Ive uploaded > the > times series ,power spectrum and the ICmap of the 0.15Hz.The upload > id is 749962 > Thanks > -Vish --Apple-Mail-18--848840480 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1 Hi

it's possible that this is = respiration, though that's be ~13sec per respiratory cycle. The cardiac = effect might show up, but at the aliased frequency, see the ssection on = respiration and cardiac effects in=A0ht= tp://www.fmrib.ox.ac.uk/analysis/techrep/tr05cb1/tr05cb1.pdf
cheers
Christian

On 10 Oct = 2007, at 19:13, Vishwadeep Ahluwalia wrote:

Hi,
I have 202 = volumes of spiral in/out data with TR=3D2s. i ran melodic on = this
data. According to the previous = mail in this topic, the highest freq that
can be = represented corresponds to 101 cycles. Ive 8 blocks i.e 8 cycles, = so
=A0=A0 =A0 = =A0 the 8th value in the text file should have a large value for = the
component=A0 =A0 =A0 =A0 corresponding to = my expt., which i do get in many
components. = However in=A0 =A0 =A0 =A0 = atleast 2 components( this is true in other
datasets as well) i get a peak=A0 =A0 =A0 =A0 at = 0.15hz(~60cycles in 404s). Is this
the peak = corresponding to respiration?=A0 =A0= =A0 =A0 Also, since my sampling
=A0 =A0 =A0 =A0 404s), and = the
cardiac pulsation being much = more than that, will i see the=A0 = =A0 =A0 =A0 peak at
the aliased = frequency or will it not be shown at all?=A0 Ive uploaded the
times series ,power spectrum and the ICmap of the = 0.15Hz.The upload id is 749962
=A0Thanks
-Vish

= --Apple-Mail-18--848840480-- ========================================================================= Date: Fri, 12 Oct 2007 17:37:21 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Duncan Mortimer <[log in to unmask]> Subject: Re: running read_avw commands in matlab In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit The solution to using the Matlab libraries on Mac OS X under bash is documented on the Mac install page on our site - did this not work? (you have to edit your startup.m file) I don't want to have to force tcsh on you :-) Duncan On 12 Oct 2007, at 16:37, Ted Yanagihara wrote: > Hi Duncan and the list, > > Our systems administrator has found a solution, or perhaps a > workaround, to the problem. I am using mac OSX that runs bash as > the default. He changed this default in Netinfo Manager so that I > startup in tcsh, then added the necessary information to keep > running fsl commands from my bash profile into the tcsh file. Now > when I check the environment that Matlab is running in using the > command you suggested, getenv('FSLDIR'), it returns the appropriate > "/usr/local/fsl" that I was looking for. > > I hope this helps anyone else who may run into this problem. I know > of at least one other person who has struggled with this on a > different system. > > thanks for your help, > > ted > > On 10/12/07, Duncan Mortimer <[log in to unmask]> wrote: > What OS are you using? > The problem is that the environment variable FSLDIR isn't being set > in the Matlab shell. > BASH is the recommended shell, and isn't the problem. Could you > issue the following in a fresh BASH shell: > > echo $FSLDIR > > If this doesn't print out the location of your FSL installation > then this is the problem, and you need to go and check the install > documentation for setting up this variable. > > If it returns the expected directory, launch Matlab and check that > the FSLDIR environment variable is still set: > > fsldir=getenv('FSLDIR') > > Duncan > > On 11 Oct 2007, at 16:39, Ted Yanagihara wrote: > >> Hello list, >> >> I am trying to read some of my images from FEAT and probtrackx in >> matlab. I have been getting an error when using read_avw and >> read_avw_complex, but the other commands (such as read_avw_hdr) >> work fine. >> >> After looking at the scripts for these functions, it appears that >> matlab cannot find the path for fsl.sh because it wants to use the >> script says to use this path: ${FSLDIR}/etc/fslconf/fsl.sh but >> matlab reads this as: /etc/fslconf/fsl.sh >> >> Is this problem because matlab is running from bash as its default >> when I need it to run in tcsh? If so, does anyone know how I can >> make my default environment tcsh? If not, what else might be going >> on? >> >> >> Here is the full error that matlab gives me when I run this set of >> commands: >> >> >> dir='~/Lab/Data/AKL_13/dti_FA'; >> >> fid=fopen(dir); >> >> image=read_avw(fid) >> Warning: The argument for the %s format specifier must be of type >> char (a string). >> > In read_avw at 30 >> sh: line 1: /etc/fslconf/fsl.sh: No such file or directory >> ??? Error using ==> fread >> Invalid file identifier. Use fopen to generate a valid file >> identifier. >> >> Error in ==> read_avw_hdr at 17 >> testval = fread(fid,1,'int32'); >> >> Error in ==> read_avw at 33 >> [dims,scales,bpp,endian,datatype]= read_avw_hdr(tmpname); > > > WWW: http://www.fmrib.ox.ac.uk/~duncan email: [log in to unmask] > > > > -- Duncan A B Mortimer DPhil MChem Computing Officer, FMRIB Centre, University of Oxford, John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK. Tel: (0)1865 222713 WWW: http://www.fmrib.ox.ac.uk/~duncan email: [log in to unmask] ========================================================================= Date: Fri, 12 Oct 2007 12:42:06 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Zhang, Xiaochu (NIH/NIDA) [F]" <[log in to unmask]> Subject: Re: RE : Re: [FSL] RE : Re: [FSL] RE : [FSL] hi and for VBM problem. In-Reply-To: A<[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: base64 VGhhbmsgeW91IHZlcnkgbXVjaCEgSW4gb3VyIGxhYiwgdGhlcmUgaXMgYSBzZXJ2ZXIgd2l0aCAx NiBDUFVzICh0aGUgYnJhbmQgaXMgSFApLiBJIGFsd2F5cyBvcGVuIGEgdGVybWluYWwgaW4gbXkg Y29tcHV0ZXIgYW5kIGRvIHNvbWUgYW5hbHlzaXMuIEhvd2V2ZXIsIG15IGNvbXB1dGVyIHdhcyBy ZWJvb3RlZCBzdWRkZW5seSBmb3Igc29tZSByZWFzb24uIFRoZXJlZm9yZSwgSSBoYXZlIHRvIHJl LW9wZW4gYSBuZXcgdGVybWluYWwgYW5kIHJlLXR5cGUgdGhlICJmc2x2Ym1fMyIgYWdhaW4uIElz IHlvdXIgbWVhbnMgdGhhdCB0aGUgImZzbHZibV8zIiB3YXMgcmVzdGFydD8gQW5kIHRoZSB0aW1l IHdhcyBiZWZvcmUgcmVib290IGlzIG5vIHVzZWZ1bCBhbnkgbW9yZT8NClRoYW5rcywgYWdhaW4h DQp4aWFvY2h1DQotLS0tLU9yaWdpbmFsIE1lc3NhZ2UtLS0tLQ0KRnJvbTogR3dlbmHDq2xsZSBE T1VBVUQgW21haWx0bzpnd2VuYWVsbGVkb3VhdWRAWUFIT08uRlJdIA0KU2VudDogMjAwN+W5tDEw 5pyIMTHml6UgMTI6MzANClRvOiBGU0xASklTQ01BSUwuQUMuVUsNClN1YmplY3Q6IFtGU0xdIFJF IDogUmU6IFtGU0xdIFJFIDogUmU6IFtGU0xdIFJFIDogW0ZTTF0gaGkgYW5kIGZvciBWQk0gcHJv YmxlbS4NCg0KSWYgeW91ciBsYWIgaGFzIGluZGVlZCBhIGNsdXN0ZXIgKHNlZSB5b3VyIElUIGd1 eXMpLA0KeW91IGNvdWxkIGFsd2F5cyByZXN0YXJ0IGl0LCBpdCB3b3VsZCBvbmx5IG92ZXJ3cml0 ZQ0Kd2hhdCBoYXMgYmVlbiBhbHJlYWR5IGNyZWF0ZWQuIEkgZ3Vlc3MgeW91IGNhbiBzdG9wIGl0 DQpieSBhIHNpbXBsZSBjdHJsIGMgaW4gdGhlIHNoZWxsIHdoZXJlIHlvdSd2ZSB0eXBlZCB0aGUN CmZzbHZibTMgY29tbWFuZC4gQnV0IEkgd291bGQgc3VnZ2VzdCB5b3Ugd2FpdCBhIGJpdCwNCmFz IGl0IHNob3VsZCBub3QgYmUgdG9vIGxvbmcgdG8gZ2l2ZSB5b3UgdGhlIHJlc3VsdHMNCm5vdy4g DQoNCkd3ZW5hZWxsZQ0KDQotLS0gIlpoYW5nLCBYaWFvY2h1IChOSUgvTklEQSkgW0ZdIg0KPHpo YW5neDJASU5UUkEuTklEQS5OSUguR09WPiBhIMOpY3JpdCA6DQoNCj4gVGhhbmsgeW91IHZlcnkg bXVjaCEgDQo+IFNvIG15IHF1ZXN0aW9uIGlzIHdoZXRoZXIgSSBhbSBhYmxlIHRvIGludGVycnVw dCB0aGUNCj4gcHJvZ3JhbSB3aGVuIGl0IGlzIHJ1bm5pbmcgKEZvciBpdCB0b29rIGEgbG9uZyB0 aW1lLA0KPiBzb21ldGltZXMgaXQgY2FuIG5vdCBiZSBhdm9pZC4gKSBhbmQgcmVzdGFydCBpdA0K PiBhZ2Fpbi4NCj4gRG9lcyBpdCBpbmZsdWVuY2UgdGhlIGZpbmFsIHJlc3VsdD8NCj4gVGhhbmtz LCBhZ2FpbiENCj4gDQo+IC0tLS0tT3JpZ2luYWwgTWVzc2FnZS0tLS0tDQo+IEZyb206IEd3ZW5h w4PCq2xsZSBET1VBVUQNCj4gW21haWx0bzpnd2VuYWVsbGVkb3VhdWRAWUFIT08uRlJdIA0KPiBT ZW50OiAyMDA3w6XCucK0MTDDpsWTy4YxMcOm4oCUwqUgMTA6MzYNCj4gVG86IEZTTEBKSVNDTUFJ TC5BQy5VSw0KPiBTdWJqZWN0OiBbRlNMXSBSRSA6IFJlOiBbRlNMXSBSRSA6IFtGU0xdIGhpIGFu ZCBmb3INCj4gVkJNIHByb2JsZW0uDQo+IA0KPiBIaSwNCj4gDQo+IHNvcnJ5LCBJIHNob3VsZCBo YXZlIGJlZW4gbW9yZSBzcGVjaWZpYy4uLiBJdCBjYW4gYmUNCj4gcnVuIGluIHBhcmFsbGVsIGlm IHlvdXIgbGFiIGhhcyBnb3QgYSBTR0Ugc3lzdGVtIG9yDQo+IHNvbWV0aGluZyBsaWtlIHRoYXQg KHNvcnJ5IEknbSBub3QgdmVyeSBnb29kIGF0DQo+IHRoZXNlDQo+IHRoaW5ncykgYW5kIHRoZW4g eW91IHNlbmQgeW91ciBqb2IgdG8gYSBidW5jaCBvZg0KPiBjb21wdXRlcnMgaW52b2x2ZWQgaW4g dGhlIGNsdXN0ZXIuIEl0IHNob3VsZCBub3QNCj4gdGFrZQ0KPiBtdWNoIGxvbmdlciBhbnl3YXku IA0KPiBIb3BlZnVsbHksIHdlIHNob3VsZCBnZXQgYSBiZXR0ZXIgYW5kIG11Y2ggcXVpY2tlcg0K PiBub24tbGluZWFyIHJlZ2lzdHJhdGlvbiAoRk5JUlQpIHNvb25pc2ggYW5kIHRoYXQNCj4gd2ls bA0KPiBiZSBpbmNsdWRlZCBpbiB0aGUgbmV4dCBGU0w0LjEgcGF0Y2guDQo+IA0KPiBDaGVlcnMs DQo+IEd3ZW5hZWxsZQ0KPiANCj4gIA0KPiAtLS0gIlpoYW5nLCBYaWFvY2h1IChOSUgvTklEQSkg W0ZdIg0KPiA8emhhbmd4MkBJTlRSQS5OSURBLk5JSC5HT1Y+IGEgw4PCqWNyaXQgOg0KPiANCj4g PiBUaGFuayB5b3UgdmVyeSBtdWNoIGZvciBzcGVlZHkgcmVzcG9uc2UhDQo+ID4gSSB0aGluayB0 aGUgZmlyc3QgcXVlc3Rpb24gaGFzIGJlZW4gcmVzb2x2ZWQuIEZvcg0KPiAyDQo+ID4gaG91cnMg cGVyIHN1YmplY3QsIGl0IHdpbGwgdGFrZSBhYm91dCAxMDAgaG91cnMNCj4gPiAoYWJvdXQgNCBk YXlzKSB0byBmaW5pc2ggdGhlIHdob2xlIDQ4IHN1YmplY3RzLiBJDQo+ID4gbmVlZCB3YWl0IHR3 byBkYXlzIG1vcmUuDQo+ID4gRm9yIHlvdSBzYWlkIGl0IGNhbiBiZSBydW4gaW4gcGFyYWxsZWwg YW5kIHRoZQ0KPiA+IGNvbXB1dGVyIGluIG91ciBsYWIgaGF2ZSBtdWx0aXBsZSBDUFVzLiBTbywg Y2FuIEkNCj4gPiBvcGVuIG11bHRpcGxlIHRlcm1pbmFscyBhbmQgcnVuIHRoZSBzY3JpcHQgaW4g ZWFjaA0KPiA+IG9uZSB0byBzaG9ydGVuIHRoZSB3aG9sZSB0aW1lPw0KPiA+IFRoYW5rcyBhZ2Fp biENCj4gPiANCj4gPiAtLS0tLU9yaWdpbmFsIE1lc3NhZ2UtLS0tLQ0KPiA+IEZyb206IEd3ZW5h w4PGksOCwqtsbGUgRE9VQVVEDQo+ID4gW21haWx0bzpnd2VuYWVsbGVkb3VhdWRAWUFIT08uRlJd IA0KPiA+IFNlbnQ6IDIwMDfDg8Klw4LCucOCwrQxMMODwqbDheKAnMOL4oCgMTHDg8Kmw6Ligqzi gJ3DgsKlIDk6MzkNCj4gPiBUbzogRlNMQEpJU0NNQUlMLkFDLlVLDQo+ID4gU3ViamVjdDogW0ZT TF0gUkUgOiBbRlNMXSBoaSBhbmQgZm9yIFZCTSBwcm9ibGVtLg0KPiA+IA0KPiA+IEhpLA0KPiA+ IA0KPiA+IEknbSBub3Qgc3VyZSBhYm91dCB5b3VyIGZpcnN0IHF1ZXN0aW9uLiBUaGUNCj4gPiBu b24tbGluZWFyDQo+ID4gcmVnaXN0cmF0aW9uIGNhbiB0YWtlIGEgd2hpbGUsIGxldCdzIHNheSBh dCBsZWFzdA0KPiAyDQo+ID4gaG91cnMgcGVyIHN1YmplY3QuIA0KPiA+IFdlIGhhdmUgZG9uZSB0 aGUgdGhpbmdzIHNvIHRoYXQgeW91IGNhbiBydW4gdGhlDQo+IGpvYnMNCj4gPiBpbiBwYXJhbGxl bCBpZiB5b3VyIGxhYiBoYXMgZ290IGEgY2x1c3RlciB0byBkbw0KPiBzby4NCj4gPiBJZg0KPiA+ IHRoaXMgaXMgbm90IHRoZSBjYXNlLCB5ZXMsIGl0IGNhbiBiZSBsb25nLCBhbmQNCj4gbXVjaA0K PiA+IGxvbmdlciB0aGFuIGZzbHZibTIgZm9yIHdoaWNoIHlvdSBoYXZlIGhvcGVmdWxseQ0KPiA+ IGNob3NlbiB0aGUgcmVjb21tZW5kZWQgYWZmaW5lIG9wdGlvbiB3aXRoIC1hLg0KPiA+IEp1c3Qg Y2hlY2sgdGhhdCB5b3VyIDREIGltYWdlIEdNX21vZF9tZXJnLm5paS5neg0KPiBpbg0KPiA+IHlv dXIgc3RhdHMgZGlyZWN0b3J5IGhhcyBiZWVuIGNyZWF0ZWQgYnkgdGhlIG1vc3QNCj4gPiByZWNl bnQgZnNsdmJtMyB5b3UgcmFuLCBhbmQgaWYgc28sIGp1c3QgYWJvcnQgdGhlDQo+ID4gZmlyc3Qg IGZzbHZibTMgY29tbWFuZC4gDQo+ID4gDQo+ID4gSG9wZSB0aGlzIGhlbHBzLA0KPiA+IEd3ZW5h ZWxsZQ0KPiA+IA0KPiA+IA0KPiA+IC0tLSAiWmhhbmcsIFhpYW9jaHUgKE5JSC9OSURBKSBbRl0i DQo+ID4gPHpoYW5neDJASU5UUkEuTklEQS5OSUguR09WPiBhIMODxpLDgsKpY3JpdCA6DQo+ID4g DQo+ID4gPiBIaSwNCj4gPiA+IA0KPiA+ID4gSSdtIHdvcmtpbmcgb24gdGhlIFZCTSBkYXRhLiBN eSBzYW1wbGUgaW5jbHVkZWQNCj4gNDgNCj4gPiA+IHN1YmplY3RzICgyNCBwYXRpZW50cw0KPiA+ ID4gYW5kIDI0IGNvbnRyb2xzKS4gSSBoYXZlIHR3byBxdWVzdGlvbnMgYWJvdXQgdGhlDQo+ID4g PiAiZnNsdmJtXzNfcHJvYyIuDQo+ID4gPiANCj4gPiA+IFRoZSBmaXJzdCBvbmUsIEkgcmFuIGl0 IDIgZGF5cyBhZ28uIEhvd2V2ZXIsIHVwDQo+IHRvDQo+ID4gPiBub3csIGl0IGhhdmUgTk9UDQo+ ID4gPiBmaW5pc2hlZC4gSSBjaGVja2VkIHRoZSBwcm9ncmFtIGlzIG5vdCBkaWVkLg0KPiA+ID4g SG93ZXZlciwgd2hlbiBJIHJ1bg0KPiA+ID4gImZzbHZibV8xIiBhbmQgImZzbHZibV8yIiwgaXQg dG9vayBubyBtb3JlIHRoYW4NCj4gNA0KPiA+ID4gaG91cnMuIElzIGl0IG5vcm1hbA0KPiA+ID4g cGhlbm9tZW5vbj8NCj4gPiA+IA0KPiA+ID4gVGhlIHNlY29uZCBvbmUsIGFib3V0IDEgZGF5IGFn bywgbXkgY29tcHV0ZXIgd2FzDQo+ID4gPiByZWJvb3RlZCBieSBwb3dlcg0KPiA+ID4gcHJvYmxl bS4gQWZ0ZXIgcmVib290LCBJIHJlLXJ1biB0aGUgImZzbHZibV8zIi4NCj4gSXQNCj4gPiA+IGxv b2tzIGxpa2UgT0suIE5lZWQgSQ0KPiA+ID4gdG8gZGVsZXRlZCBhbGwgZmlsZXMgYW5kIHJlc3Rh cnQgZnJvbQ0KPiAiZnNsdmJtXzEiPyANCj4gPiANCj4gPiA+IA0KPiA+ID4gVGhhbmsgeW91IHZl cnkgbXVjaCENCj4gPiA+IA0KPiA+ID4gWGlhb2NodSBaaGFuZyBQaC5EDQo+ID4gPiANCj4gPiA+ IFZpc2l0aW5nIFJlc2VhcmNoIEZlbGxvdw0KPiA+ID4gDQo+ID4gPiBOSUgvTklEQS1JUlANCj4g PiA+IA0KPiA+ID4gNTUwMCBOYXRoYW4gU2hvY2sgRHJpdmUNCj4gPiA+IA0KPiA+ID4gQmFsdGlt b3JlIE1EIDIxMjI0DQo+ID4gPiANCj4gPiA+ICANCj4gPiA+IA0KPiA+ID4gVGVsOiAoNDEwKSAt IDU1MCAtIDE0NDAgZXh0LiA0MzQNCj4gPiA+IA0KPiA+ID4gIA0KPiA+ID4gDQo+ID4gPiANCj4g PiANCj4gPiANCj4gPiANCj4gPiAgICAgIA0KPiA+DQo+DQpfX19fX19fX19fX19fX19fX19fX19f X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fXw0K PiA+IA0KPiA+IE5lIGdhcmRleiBwbHVzIHF1J3VuZSBzZXVsZSBhZHJlc3NlIG1haWwgISBDb3Bp ZXoNCj4gPiB2b3MgbWFpbHMgdmVycyBZYWhvbyEgTWFpbCANCj4gPiANCj4gDQo+IA0KPiANCj4g ICAgICANCj4NCl9fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19f X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fDQo+IA0KPiBOZSBnYXJkZXogcGx1cyBxdSd1 bmUgc2V1bGUgYWRyZXNzZSBtYWlsICEgQ29waWV6DQo+IHZvcyBtYWlscyB2ZXJzIFlhaG9vISBN YWlsIA0KPiANCg0KDQoNCiAgICAgIF9fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19f X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fIA0KTmUgZ2FyZGV6IHBs dXMgcXUndW5lIHNldWxlIGFkcmVzc2UgbWFpbCAhIENvcGlleiB2b3MgbWFpbHMgdmVycyBZYWhv byEgTWFpbCANCg== ========================================================================= Date: Fri, 12 Oct 2007 19:06:50 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Markus Gschwind <[log in to unmask]> Subject: DTI processing MIME-version: 1.0 Content-transfer-encoding: 7BIT Content-disposition: inline Content-type: text/plain; charset=ISO-8859-1; DelSp=Yes; format=flowed Hi FSLers! I am struggling with the DTI Mode of FSL: 1. I can open FA data in FSLveiw 2D nicely, but I cannot open any FA-data (produced in FSL3 on Cygwin) in FSLview 3D. That means: It atually works but I just don't see anything! Is that normal? 2. I installed FSL 4 in a vmplayer. Atotest with the testdata worked fine. After Eddy corrent-correction (30 volumes of 30 have been processed. "Done!") it tells me ERROR: window name "prompt" already exists in parent window name "prompt" already exists in parent while executing "toplevel .prompt -borderwidth 5" (procedure "MxPause" line 4) invoked from within "MxPause " Done! " " (procedure "fdt:apply" line 313) invoked from within "fdt:apply .fdt keep" invoked from within ".fdt.apply invoke" ("uplevel" body line 1) invoked from within "uplevel #0 [list $w invoke]" (procedure "tk::ButtonUp" line 22) invoked from within "tk::ButtonUp .fdt.apply" (command bound to event) and: Errors: terminate called after throwing an instance on 'Newmat::InompatibleDimensionsException'. 3. Trying Bedpostx on another data I got THE SAME error message. What is wrong? How can I fix that? Thank you very much! Cheers, Markus -- Markus Gschwind, M.D. Laboratory for Neurology and Imaging of Cognition Dept of Neurosciences University Medical Center (CMU) 1 Michel-Servet - 1211 GENEVA - CH Tel 0041 (0) 22 379 5324 Fax 0041 (0) 22 379 5402 email: [log in to unmask] http://labnic.unige.ch ========================================================================= Date: Fri, 12 Oct 2007 13:14:31 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Lokke Highstein <[log in to unmask]> Subject: FAST on OSX cluster only running on head node Mime-Version: 1.0 (Apple Message framework v752.2) Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed hello list, i'm sure you are all sick of my constant posts about the xserve cluster, but i have another question. we are successfully getting FEAT, bedpost, and probtrack jobs to run through the SGE, and get distributed across the cluster, but when we run FAST, it doesn't get submitted to the queue, and only runs on the head node. i tried reading the fast.tcl file and searching for the terms long.q all.q fsl_sub and SGE, didn't find any of them there. is there something i need to edit to make FAST jobs get submitted to my SGE queue? thanks, lokke ========================================================================= Date: Fri, 12 Oct 2007 19:41:51 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Martin Kavec <[log in to unmask]> Subject: Re: FAST on OSX cluster only running on head node In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi Lokke, I think that FAST can not be parallelized (as simply) as other FSL tools. Segmentation is performed on the whole image at once and not slice by slice. So I assume that the FSL guys didn't configure it for execution on the cluster, as it would not give us the results faster anyway. Martin On Friday 12 October 2007 19:14:31 Lokke Highstein wrote: > hello list, > > i'm sure you are all sick of my constant posts about the xserve > cluster, but i have another question. > > we are successfully getting FEAT, bedpost, and probtrack jobs to run > through the SGE, and get distributed across the cluster, but when we > run FAST, it doesn't get submitted to the queue, and only runs on the > head node. > > i tried reading the fast.tcl file and searching for the terms long.q > all.q fsl_sub and SGE, didn't find any of them there. > > is there something i need to edit to make FAST jobs get submitted to > my SGE queue? > > thanks, > > lokke ========================================================================= Date: Fri, 12 Oct 2007 19:09:53 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Rachel <[log in to unmask]> Subject: eddy_correct Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="windows-1252" Hi,=20 I have a DTI dataset for which I would like to create an index of data=20= quality for each subject (that I could regress out in correlations with=20= other measures). I was hoping to use an index derived from the output of=20= the eddy current correction program, but I am not sure how to interpret=20= the log file. What are the 4 columns and 4 rows for each volume, and what= =20 do the numbers represent? Thank you, Rachel ========================================================================= Date: Fri, 12 Oct 2007 14:28:07 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matt Glasser <[log in to unmask]> Organization: ma-tea.com Subject: Re: eddy_correct In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Rachel, Those are the transformation matrices required to align each volume to the reference volume in a 4D dataset. I don't think it would provide a good measure of data quality. Peace, Matt. -----Original Message----- From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Rachel Sent: Friday, October 12, 2007 2:10 PM To: [log in to unmask] Subject: [FSL] eddy_correct Hi, I have a DTI dataset for which I would like to create an index of data quality for each subject (that I could regress out in correlations with other measures). I was hoping to use an index derived from the output of the eddy current correction program, but I am not sure how to interpret the log file. What are the 4 columns and 4 rows for each volume, and what do the numbers represent? Thank you, Rachel ========================================================================= Date: Fri, 12 Oct 2007 15:45:04 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Lokke Highstein <[log in to unmask]> Subject: Re: FAST on OSX cluster only running on head node In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit martin, thanks for the quick reply. that makes sense, but it would still be very useful to dole the jobs out to whatever node is least loaded, or to keep the job in queue until a node becomes available. in fact this is what happens with FEAT as far as i can tell. so, i'm still hoping that there's a way to get this to go through the queue system. thanks lokke On Oct 12, 2007, at 1:41 PM, Martin Kavec wrote: > Hi Lokke, > > I think that FAST can not be parallelized (as simply) as other FSL > tools. > Segmentation is performed on the whole image at once and not slice > by slice. > So I assume that the FSL guys didn't configure it for execution on the > cluster, as it would not give us the results faster anyway. > > Martin > > On Friday 12 October 2007 19:14:31 Lokke Highstein wrote: >> hello list, >> >> i'm sure you are all sick of my constant posts about the xserve >> cluster, but i have another question. >> >> we are successfully getting FEAT, bedpost, and probtrack jobs to run >> through the SGE, and get distributed across the cluster, but when we >> run FAST, it doesn't get submitted to the queue, and only runs on the >> head node. >> >> i tried reading the fast.tcl file and searching for the terms long.q >> all.q fsl_sub and SGE, didn't find any of them there. >> >> is there something i need to edit to make FAST jobs get submitted to >> my SGE queue? >> >> thanks, >> >> lokke ========================================================================= Date: Fri, 12 Oct 2007 22:29:46 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Christopher Bell <[log in to unmask]> Subject: fsl_glm Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Hello, I am interested in further understanding the new fsl_glm. I was trying to= output the z-stats, but fsl_glm would not run without an entered value fo= r --out_z=3D? Does this value represent a threshold? Can I output unthresho= lded z-stats? Thanks in advance! Chris Bell ========================================================================= Date: Fri, 12 Oct 2007 14:39:02 -0700 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Junghee Lee <[log in to unmask]> Subject: installing problem of fsl 4.0 In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v624) Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: quoted-printable Hello Everyone, I've tried to install FSL 4.0 on Apple G5 running OS X 10.3.8. I used=20= the installer script and didn't have any problem until I tried to=20 launch fsl gui. When I tried to use GUI, I got the following error message. I even=20 tried to run FSL GUI straight from X 11 and got the same message. JQLab-G530:~/Desktop/Download jedimaster$ dyld:=20 /usr/local/fsl/extras/bin/fslwish8.4 Undefined symbols: /usr/local/fsl/extras/lib/libtcl8.4.dylib undefined reference to=20 _OSSpinLockLock expected to be defined in /usr/lib/libSystem.B.dylib I also copied the installation process below. Any feedback would be=20 greatly appreciated. Junghee JQLab-G530:~/Desktop/Download jedimaster$ sudo sh fsl_installer.sh Password: Looking for FSL tarball in the current directory... *************************************************************** No FSL tarball specified, assuming you want me to install /Users/jedimaster/Desktop/Download/fsl-4.0.1-macosx.tar from the current directory. *************************************************************** FSL instal script =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D Where would you like to install FSL to? [/usr/local] '/usr/local/fsl' exists - would you like to remove it? [yes] Deleting '/usr/local/fsl'... Installing FSL from=20 /Users/jedimaster/Desktop/Download/fsl-4.0.1-macosx.tar into=20 '/usr/local'... Untarring fsl-4.0.1-macosx.tar Would you like to install FSLView.app into /Applications (requires=20 Administrator priviledges)? [yes] Moving old FSLView into the Trash Installing FSLView into /Applications Patching environment '.bash_profile' already contains a FSLDIR definition which is correctly=20= configured, not changing. Setting up Apple terminal... DISPLAY is already being configured in your=20 '/Users/jedimaster/.bash_profile' - not changing JQLab-G530:~/Desktop/Download jedimaster$ $FSLDIR -bash: /usr/local/fsl: is a directory JQLab-G530:~/Desktop/Download jedimaster$ flirt -version FLIRT version 5.4.2 JQLab-G530:~/Desktop/Download jedimaster$ fsl & [1] 742 JQLab-G530:~/Desktop/Download jedimaster$ dyld:=20 /usr/local/fsl/extras/bin/fslwish8.4 Undefined symbols: /usr/local/fsl/extras/lib/libtcl8.4.dylib undefined reference to=20 _OSSpinLockLock expected to be defined in /usr/lib/libSystem.B.dylib [1]+ Trace/BPT trap fsl On Oct 1, 2007, at 4:35 AM, Matthew Webster wrote: > Hi, > To make X11 terminal run fsl you need to add -ls to the xterm=20 > command than X11 runs ( go to X11-Applications-Customize Menu and=20 > change xterm ) to make it read your .profile when starting a shell.. > > Matthew >> Hello, >> >> I've just installed FSL on my new Intel Mac running OS X Tiger. I=20 >> used the >> installer script and installed FSL into the /Applications directory. =20= >> When I >> run the tests using the $FSLDIR command, etc. in terminal all is = fine, >> however, when I go to X11 and type fsl I get the error that "bash:=20 >> fsl: >> command not found". I can launch fsl from within terminal by typing=20= >> fsl just >> not from the X11 terminal. >> >> Also when I open terminal I used to get the following error: >> /Applications/fsl/etc/fslconf/fslmachtype.sh: line 1: gcc: command = not >> found. I found an e-mail in the archive which advised editing the >> fslmachtype.sh file which I have done. I now get a new error that=20 >> says: >> /Applications/fsl/etc/fslconf/fslmachtype.sh: line 27: [0: command=20 >> not found. >> >> I am not sure if the above two problems are connected or what I=20 >> should do >> about either of them. Thank in advance for the help. >> >> Liz > > ------------------------------------ Junghee Lee Post-doctoral fellow Semel Institute of Neuroscience and Human Behavior, UCLA UCLA/VA Greater Los Angeles Healthcare System 11301 Wilshire Blvd. MIRECC, Bldg. 210=A0 (MC 210A) Los Angeles, CA 90073 310-478-3711 ext.42986 http://greenlab.npih.ucla.edu ========================================================================= Date: Fri, 12 Oct 2007 22:42:42 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Christian F. Beckmann" <[log in to unmask]> Subject: Re: fsl_glm In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: multipart/alternative; boundary=Apple-Mail-19--830202124 --Apple-Mail-19--830202124 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Hi, use --out_z=blah to save the unthresholded z-stats in a file 'blah' regards Christian _______________________________________________ Christian F. Beckmann, DPhil Senior Lecturer, Clinical Neuroscience Department Division of Neuroscience and Mental Health Imperial College London Hammersmith Campus, Rm 247, Cyclotron Bldg Du Cane Road, London W12 0NN, UK Tel.: +44 (0) 208 383 8598 --- Fax: +44 (0) 208 383 2029 Email: [log in to unmask] http://www.imperial.ac.uk/medicine/people/c.beckmann/ On 12 Oct 2007, at 22:29, Christopher Bell wrote: > Hello, > > I am interested in further understanding the new fsl_glm. I was > trying to > output the z-stats, but fsl_glm would not run without an entered > value for > --out_z=? Does this value represent a threshold? Can I output > unthresholded > z-stats? Thanks in advance! > > Chris Bell --Apple-Mail-19--830202124 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1 Hi,

use --out_z=3Dblah=A0= to save the unthresholded z-stats in a file = 'blah'
Christian

_____________________________= __________________
Christian F. Beckmann, = DPhil
Senior Lecturer,=A0Clinical Neuroscience = Department
Division of Neuroscience and Mental = Health
Imperial College London
Hammersmith = Campus,=A0Rm 247, Cyclotron Bldg
Du Cane Road, London W12 0NN, = UK
Tel.: +44 (0) 208 383 8598=A0 =A0---=A0 =A0Fax: +44 (0) 208 = 383 2029

Email:= [log in to unmask]=A0= =A0



On 12 = Oct 2007, at 22:29, Christopher Bell wrote:

Hello,

I am interested in further = understanding the new fsl_glm. I was trying to
output the z-stats, but fsl_glm would not run = without an entered value for
--out_z=3D? = Does this value represent a threshold? Can I output = unthresholded
z-stats? Thanks in = advance!

Chris Bell
=

= --Apple-Mail-19--830202124-- ========================================================================= Date: Fri, 12 Oct 2007 16:44:47 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Ted Yanagihara <[log in to unmask]> Subject: Re: running read_avw commands in matlab In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_24383_17990071.1192221887537" ------=_Part_24383_17990071.1192221887537 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi Duncan, I actually was unable to find the startup.m file (not just in the matlab folder, but anywhere on my computer), so I did not set the environment variables up. Rather that waste people's time with it, I'm going to just hope that the workaround keeps things swept under the rug. I really appreciate the attention you've given my issue, thanks again! ted On 10/12/07, Duncan Mortimer <[log in to unmask]> wrote: > > The solution to using the Matlab libraries on Mac OS X under bash is > documented on the Mac install page on our site - did this not work? > (you have to edit your startup.m file) > > I don't want to have to force tcsh on you :-) > > Duncan > > > On 12 Oct 2007, at 16:37, Ted Yanagihara wrote: > > > Hi Duncan and the list, > > > > Our systems administrator has found a solution, or perhaps a > > workaround, to the problem. I am using mac OSX that runs bash as > > the default. He changed this default in Netinfo Manager so that I > > startup in tcsh, then added the necessary information to keep > > running fsl commands from my bash profile into the tcsh file. Now > > when I check the environment that Matlab is running in using the > > command you suggested, getenv('FSLDIR'), it returns the appropriate > > "/usr/local/fsl" that I was looking for. > > > > I hope this helps anyone else who may run into this problem. I know > > of at least one other person who has struggled with this on a > > different system. > > > > thanks for your help, > > > > ted > > > > On 10/12/07, Duncan Mortimer <[log in to unmask]> wrote: > > What OS are you using? > > The problem is that the environment variable FSLDIR isn't being set > > in the Matlab shell. > > BASH is the recommended shell, and isn't the problem. Could you > > issue the following in a fresh BASH shell: > > > > echo $FSLDIR > > > > If this doesn't print out the location of your FSL installation > > then this is the problem, and you need to go and check the install > > documentation for setting up this variable. > > > > If it returns the expected directory, launch Matlab and check that > > the FSLDIR environment variable is still set: > > > > fsldir=getenv('FSLDIR') > > > > Duncan > > > > On 11 Oct 2007, at 16:39, Ted Yanagihara wrote: > > > >> Hello list, > >> > >> I am trying to read some of my images from FEAT and probtrackx in > >> matlab. I have been getting an error when using read_avw and > >> read_avw_complex, but the other commands (such as read_avw_hdr) > >> work fine. > >> > >> After looking at the scripts for these functions, it appears that > >> matlab cannot find the path for fsl.sh because it wants to use the > >> script says to use this path: ${FSLDIR}/etc/fslconf/fsl.sh but > >> matlab reads this as: /etc/fslconf/fsl.sh > >> > >> Is this problem because matlab is running from bash as its default > >> when I need it to run in tcsh? If so, does anyone know how I can > >> make my default environment tcsh? If not, what else might be going > >> on? > >> > >> > >> Here is the full error that matlab gives me when I run this set of > >> commands: > >> > >> >> dir='~/Lab/Data/AKL_13/dti_FA'; > >> >> fid=fopen(dir); > >> >> image=read_avw(fid) > >> Warning: The argument for the %s format specifier must be of type > >> char (a string). > >> > In read_avw at 30 > >> sh: line 1: /etc/fslconf/fsl.sh: No such file or directory > >> ??? Error using ==> fread > >> Invalid file identifier. Use fopen to generate a valid file > >> identifier. > >> > >> Error in ==> read_avw_hdr at 17 > >> testval = fread(fid,1,'int32'); > >> > >> Error in ==> read_avw at 33 > >> [dims,scales,bpp,endian,datatype]= read_avw_hdr(tmpname); > > > > > > WWW: http://www.fmrib.ox.ac.uk/~duncan email: [log in to unmask] > > > > > > > > > > -- > Duncan A B Mortimer DPhil MChem > Computing Officer, FMRIB Centre, University of Oxford, > John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK. > Tel: (0)1865 222713 > WWW: http://www.fmrib.ox.ac.uk/~duncan email: [log in to unmask] > ------=_Part_24383_17990071.1192221887537 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi Duncan,

I actually was unable to find the startup.m file (not just in the matlab folder, but anywhere on my computer), so I did not set the environment variables up. Rather that waste people's time with it, I'm going to just hope that the workaround keeps things swept under the rug.

I really appreciate the attention you've given my issue, thanks again!

ted


On 10/12/07, Duncan Mortimer < [log in to unmask]> wrote:
The solution to using the Matlab libraries on Mac OS X under bash is
documented on the Mac install page on our site - did this not work?
(you have to edit your startup.m file)

I don't want to have to force tcsh on you :-)

Duncan


On 12 Oct 2007, at 16:37, Ted Yanagihara wrote:

> Hi Duncan and the list,
>
> Our systems administrator has found a solution, or perhaps a
> workaround, to the problem. I am using mac OSX that runs bash as
> the default. He changed this default in Netinfo Manager so that I
> startup in tcsh, then added the necessary information to keep
> running fsl commands from my bash profile into the tcsh file. Now
> when I check the environment that Matlab is running in using the
> command you suggested, getenv('FSLDIR'), it returns the appropriate
> "/usr/local/fsl" that I was looking for.
>
> I hope this helps anyone else who may run into this problem. I know
> of at least one other person who has struggled with this on a
> different system.
>
> thanks for your help,
>
> ted
>
> On 10/12/07, Duncan Mortimer < [log in to unmask]> wrote:
> What OS are you using?
> The problem is that the environment variable FSLDIR isn't being set
> in the Matlab shell.
> BASH is the recommended shell, and isn't the problem. Could you
> issue the following in a fresh BASH shell:
>
> echo $FSLDIR
>
> If this doesn't print out the location of your FSL installation
> then this is the problem, and you need to go and check the install
> documentation for setting up this variable.
>
> If it returns the expected directory, launch Matlab and check that
> the FSLDIR environment variable is still set:
>
> fsldir=getenv('FSLDIR')
>
> Duncan
>
> On 11 Oct 2007, at 16:39, Ted Yanagihara wrote:
>
>> Hello list,
>>
>> I am trying to read some of my images from FEAT and probtrackx in
>> matlab. I have been getting an error when using read_avw and
>> read_avw_complex, but the other commands (such as read_avw_hdr)
>> work fine.
>>
>> After looking at the scripts for these functions, it appears that
>> matlab cannot find the path for fsl.sh because it wants to use the
>> script says to use this path: ${FSLDIR}/etc/fslconf/fsl.sh   but
>> matlab reads this as: /etc/fslconf/fsl.sh
>>
>> Is this problem because matlab is running from bash as its default
>> when I need it to run in tcsh? If so, does anyone know how I can
>> make my default environment tcsh? If not, what else might be going
>> on?
>>
>>
>> Here is the full error that matlab gives me when I run this set of
>> commands:
>>
>> >> dir='~/Lab/Data/AKL_13/dti_FA';
>> >> fid=fopen(dir);
>> >> image=read_avw(fid)
>> Warning: The argument for the %s format specifier must be of type
>> char (a string).
>> > In read_avw at 30
>> sh: line 1: /etc/fslconf/fsl.sh: No such file or directory
>> ??? Error using ==> fread
>> Invalid file identifier.  Use fopen to generate a valid file
>> identifier.
>>
>> Error in ==> read_avw_hdr at 17
>> testval = fread(fid,1,'int32');
>>
>> Error in ==> read_avw at 33
>>   [dims,scales,bpp,endian,datatype]= read_avw_hdr(tmpname);
>
>
> WWW: http://www.fmrib.ox.ac.uk/~duncan    email: [log in to unmask]
>
>
>
>

--
Duncan A B Mortimer DPhil MChem
                 Computing Officer, FMRIB Centre, University of Oxford,
               John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK.
Tel: (0)1865 222713
WWW: http://www.fmrib.ox.ac.uk/~duncan    email: [log in to unmask]

------=_Part_24383_17990071.1192221887537-- ========================================================================= Date: Fri, 12 Oct 2007 20:52:32 -0700 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Jeremy Bronson <[log in to unmask]> Subject: Re: FSL on XGrid w/ XServe (now SGE) Content-Type: text/plain; charset="us-ascii" Content-Disposition: inline Content-Transfer-Encoding: 7bit MIME-Version: 1.0 Hi Lokke, Thanks for the response. We have a very similar setup, and as usual all users authenticate via LDAP/Kerberos to the XServe for access to their accounts, shares and Kerberized services. This works for Windows and Linux clients as well as Mac. The problem is that SGE seems to explicitly state that the nodes in the grid must have identical _local_ accounts. This makes a certain kind of sense, it's the other way to solve the issue of permissions besides using the 'nobody' user, but it doesn't seem like it scales well. Imagine trying to administer a grid at a university or company across buildings, networks, os types and versions, and ensuring that every system has the same user accounts with the same IDs per user - it's functionally impossible. The other method using 'nobody' is used by XGrid, which allows for massively distributed grid processing, a la SETI@Home and Stanford's XGrid projects. Any user anywhere in the world can elect to become a grid node without any special user accounts on their system. It makes me think that I've perhaps misunderstood SGE's requirements, so I was hoping to get input from someone who has it fully implemented and functioning. Steve? Thanks again! Jeremy ----- Original Message ---- From: Lokke Highstein <[log in to unmask]> To: [log in to unmask] Sent: Tuesday, October 9, 2007 4:31:50 PM Subject: Re: [FSL] FSL on XGrid w/ XServe (now SGE) hi jeremy, i'm very new to this but i can already answer one of your questions and then defer to the others here on the second two. we have a 8 node xserve cluster and we use open directory as our master username/password database management system. just about anything that can access LDAP data can authenticate to it so most of the computers (servers included) can share the same username/password database (although the windows xp boxes give me trouble.) open directory is built into OSX server, and on the xserve nodes it's very simple to set up so they authenticate to that system. i'm not sure about permissions, i am still playing with them at the moment, and i also am just starting to look into multiple queues. i'm sure someone else will have more info. -lokke On Oct 9, 2007, at 3:17 PM, Jeremy Bronson wrote: > Hi, > > I tried to get FSL working with XGrid during the summer, but wasn't > quite successful, and was recommended to use SGE in its place by > Steve and > others. I'm giving it another go, but I've run into a few questions > while setting up SGE. Hopefully they're easy enough to answer... > > -The SGE manual says that each host must have the same user account > names and passwords, which is the alternate to XGrid's nobody:nobody > permissions, but seems equally impractical. If SGE is really > designed to be > implemented in large grids, I'm not sure how it could scale beyond a > handful of managed machines. My question is whether this is how the > lab is configured at Oxford or any other site, or alternately are all > jobs submitted under a single username? > > -All scan data resides on an XServe, automounted via NFS. What kind of > permissions are necessary on this share? > > -Is anyone using the multiple queues available in SGE, or only using > one for large jobs? > > My basic goal is to allow users to login, start multiple large jobs on > the grid, then log out and retrieve the results later. The FEAT > first-level analyses tend to tie up all the machines for a time, > so I'd love > to speed them up by running the jobs in parallel and in the > background. > > > Thanks in advance! > > Jeremy > > > > > > > On Tue, 14 Aug 2007 06:24:00 +0100, Steve Smith <[log in to unmask]> > wrote: >> Hi Jeremy, >> >> I agree with Andrew, most people seem much happier with SGE than with > >> XGrid, so if it's not too late I would consider that. >> >> Anyway - yes, hopefully FSL 4.0 should be fairly easy to setup with >> either (though much easier with SGE as that's what we have so should > >> need much less customising). >> >> First, see the brief intro at: >> http://www.fmrib.ox.ac.uk/fsl/fsl/downloading.html#sge >> So if the user runs any of these programs on a machine which can >> submit to the cluster then that should happen automatically, after >> the sysadmin has: >> >> Setup the central cluster-controlling script >> $FSLDIR/bin/fsl_sub >> >> This is a heavily commented shell script and hopefully should be >> reasonably easy to follow and customise..... >> >> >> So hopefully the user can just run FSL programs on their normal >> working machine, and if it's setup to be a cluster submission host, >> then whenever a "big" job is run that will automatically get sent to > >> the queue. >> >> Cheers, Steve. >> >> >> >> >> >> On 14 Aug 2007, at 05:59, Jeremy Bronson wrote: >> >>> Hi All, >>> >>> I'm the new administrator of a neuroimaging research lab, and I'm >>> working on getting FSL and other MRI-analysis tools running on >>> XGrid. I'm not yet intimately familiar with how FSL works, so I'm > >>> hoping there are others out there who have already figured out how > >>> to run it on a cluster, so I don't have to reinvent the wheel. >>> I've heard of several existing clusters that run FSL, albeit a >>> modified version. I'm hoping it can be done with the off-the-shelf > >>> version, perhaps it's now possible with the just-released 4.0? If > >>> anybody could point me in the direction of some specifics, I'd be >>> most grateful. >>> >>> We've got an XServe with RAID that houses all the data, and FSL is > >>> installed and configured on all host (agent) machines. Most users > >>> use the GUI, and manually point the tools to the appropriate data, > >>> so I assume they'll need to familiarize themselves with the > command- >>> line tools and specifying data directories on the CLI (or via >>> GridStuffer). I'm thinking that each machine will need the data >>> volume to be auto-mounted (NFS?) at startup with appropriate >>> directories having read/write access for the 'nobody' user. Does >>> this sound correct? >>> >>> Additionally, each tool that's part of the FSL package seems to >>> launch a number of other UNIX commands during analysis, to copy, >>> move and otherwise manipulate the result data. Will this confuse >>> XGrid, or will the job and all sub-commands run until the original > >>> command completes? (e.g. A complete FEAT analysis) >>> >>> Hopefully this isn't too difficult, and afterwards I'd like to take > >>> the time to post a HOWTO on macresearch.org or the like, so others > >>> might take advantage of the info. Thanks in advance to anyone who > >>> might be able to help. >>> >>> >>> Jeremy Bronson >>> Systems Administrator >>> Frey Research Lab >>> University of Oregon >> >> >> > > ---------------------------------------------------------------------- > -- >> --- >> Stephen M. Smith, Professor of Biomedical Engineering >> Associate Director, Oxford University FMRIB Centre >> >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >> +44 (0) 1865 222726 (fax 222717) >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >> > > ---------------------------------------------------------------------- > -- >> --- >> > > > > > > Hi, > > I tried to get FSL working with XGrid during the summer, but wasn't > quite successful, and was recommended to use SGE in its place by > Steve and > others. I'm giving it another go, but I've run into a few questions > while setting up SGE. Hopefully they're easy enough to answer... > > -The SGE manual says that each host must have the same user account > names and passwords, which is the alternate to XGrid's nobody:nobody > permissions, but seems equally impractical. If SGE is really > designed to be > implemented in large grids, I'm not sure how it could scale beyond a > handful of managed machines. My question is whether this is how the > lab is configured at Oxford or any other site, or alternately are all > jobs submitted under a single username? > > -All scan data resides on an XServe, automounted via NFS. What kind of > permissions are necessary on this share? > > -Is anyone using the multiple queues available in SGE, or only using > one for large jobs? > > My basic goal is to allow users to login, start multiple large jobs on > the grid, then log out and retrieve the results later. The FEAT > first-level analyses tend to tie up all the machines for a time, > so I'd love > to speed them up by running the jobs in parallel and in the > background. > > > Thanks in advance! > > Jeremy > > > > > > > On Tue, 14 Aug 2007 06:24:00 +0100, Steve Smith <[log in to unmask]> > wrote: >> Hi Jeremy, >> >> I agree with Andrew, most people seem much happier with SGE than with > >> XGrid, so if it's not too late I would consider that. >> >> Anyway - yes, hopefully FSL 4.0 should be fairly easy to setup with >> either (though much easier with SGE as that's what we have so should > >> need much less customising). >> >> First, see the brief intro at: >> http://www.fmrib.ox.ac.uk/fsl/fsl/downloading.html#sge >> So if the user runs any of these programs on a machine which can >> submit to the cluster then that should happen automatically, after >> the sysadmin has: >> >> Setup the central cluster-controlling script >> $FSLDIR/bin/fsl_sub >> >> This is a heavily commented shell script and hopefully should be >> reasonably easy to follow and customise..... >> >> >> So hopefully the user can just run FSL programs on their normal >> working machine, and if it's setup to be a cluster submission host, >> then whenever a "big" job is run that will automatically get sent to > >> the queue. >> >> Cheers, Steve. >> >> >> >> >> >> On 14 Aug 2007, at 05:59, Jeremy Bronson wrote: >> >>> Hi All, >>> >>> I'm the new administrator of a neuroimaging research lab, and I'm >>> working on getting FSL and other MRI-analysis tools running on >>> XGrid. I'm not yet intimately familiar with how FSL works, so I'm > >>> hoping there are others out there who have already figured out how > >>> to run it on a cluster, so I don't have to reinvent the wheel. >>> I've heard of several existing clusters that run FSL, albeit a >>> modified version. I'm hoping it can be done with the off-the-shelf > >>> version, perhaps it's now possible with the just-released 4.0? If > >>> anybody could point me in the direction of some specifics, I'd be >>> most grateful. >>> >>> We've got an XServe with RAID that houses all the data, and FSL is > >>> installed and configured on all host (agent) machines. Most users > >>> use the GUI, and manually point the tools to the appropriate data, > >>> so I assume they'll need to familiarize themselves with the > command- >>> line tools and specifying data directories on the CLI (or via >>> GridStuffer). I'm thinking that each machine will need the data >>> volume to be auto-mounted (NFS?) at startup with appropriate >>> directories having read/write access for the 'nobody' user. Does >>> this sound correct? >>> >>> Additionally, each tool that's part of the FSL package seems to >>> launch a number of other UNIX commands during analysis, to copy, >>> move and otherwise manipulate the result data. Will this confuse >>> XGrid, or will the job and all sub-commands run until the original > >>> command completes? (e.g. A complete FEAT analysis) >>> >>> Hopefully this isn't too difficult, and afterwards I'd like to take > >>> the time to post a HOWTO on macresearch.org or the like, so others > >>> might take advantage of the info. Thanks in advance to anyone who > >>> might be able to help. >>> >>> >>> Jeremy Bronson >>> Systems Administrator >>> Frey Research Lab >>> University of Oregon >> >> >> > > ---------------------------------------------------------------------- > -- >> --- >> Stephen M. Smith, Professor of Biomedical Engineering >> Associate Director, Oxford University FMRIB Centre >> >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >> +44 (0) 1865 222726 (fax 222717) >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >> > > ---------------------------------------------------------------------- > -- >> --- >> > > > > > > ========================================================================= Date: Sat, 13 Oct 2007 08:54:13 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: FAST on OSX cluster only running on head node In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - Martin was right - you won't get parallelisation on FAST but there's no reason not to use the queues to run a job. You can just prepend the calls to the fast/mfast programs in $FSLDIR/tcl/ fast.tcl ; find: if { $channels == 1 } { set thecommand "$FSLDIR/bin/fast -t$type" } else { set thecommand "$FSLDIR/bin/mfast -s $channels" } And put the appropriate call to $FSLDIR/bin/fsl_sub -T 15 (for example) at the beginning of the commands. Alternatively, if users are using command line programs, they can just put the same command before the rest of the command. Cheers. On 12 Oct 2007, at 20:45, Lokke Highstein wrote: > martin, > > thanks for the quick reply. > > that makes sense, but it would still be very useful to dole the > jobs out to whatever node is least loaded, or to keep the job in > queue until a node becomes available. in fact this is what happens > with FEAT as far as i can tell. > > so, i'm still hoping that there's a way to get this to go through > the queue system. > > thanks > > lokke > On Oct 12, 2007, at 1:41 PM, Martin Kavec wrote: > >> Hi Lokke, >> >> I think that FAST can not be parallelized (as simply) as other FSL >> tools. >> Segmentation is performed on the whole image at once and not slice >> by slice. >> So I assume that the FSL guys didn't configure it for execution on >> the >> cluster, as it would not give us the results faster anyway. >> >> Martin >> >> On Friday 12 October 2007 19:14:31 Lokke Highstein wrote: >>> hello list, >>> >>> i'm sure you are all sick of my constant posts about the xserve >>> cluster, but i have another question. >>> >>> we are successfully getting FEAT, bedpost, and probtrack jobs to run >>> through the SGE, and get distributed across the cluster, but when we >>> run FAST, it doesn't get submitted to the queue, and only runs on >>> the >>> head node. >>> >>> i tried reading the fast.tcl file and searching for the terms long.q >>> all.q fsl_sub and SGE, didn't find any of them there. >>> >>> is there something i need to edit to make FAST jobs get submitted to >>> my SGE queue? >>> >>> thanks, >>> >>> lokke ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Sat, 13 Oct 2007 12:43:17 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Emily RUBIN-FERREIRA <[log in to unmask]> Subject: fslvm32 In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi, I would like to install FSL 4.0 on a Windows XP computer 64-bit, using vmware and fslvm32 as instructed. I was reading in the FAQ the following: that on 32-bit computers, there is a limit on the size that a single program can be regardless of the RAM/swap you have, and that the limit is 2 GB. If fslvm is 32bit, does this mean that if I install using fslvm, that my computer becomes like a 32-bit computer? Thanks, emily ========================================================================= Date: Sat, 13 Oct 2007 21:26:32 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Zhang, Xiaochu (NIH/NIDA) [F]" <[log in to unmask]> Subject: The orientation problem in VBM analysis MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01C80E01.41217E8D" This is a multi-part message in MIME format. ------_=_NextPart_001_01C80E01.41217E8D Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: quoted-printable Hi experts, I am working on the VBM analysis. For the data is from multiple studies, some is sagitial and some is axial. After "tbss_1" and check the slicedir, the data showed inconsistent orientation. Who can help me and tell me how to change the orientation? Thanks a lot! =20 Xiaochu Zhang Ph.D Visiting Research Fellow NIH/NIDA-IRP 5500 Nathan Shock Drive Baltimore MD 21224 =20 Tel: (410) - 550 - 1440 ext. 434 =20 ------_=_NextPart_001_01C80E01.41217E8D Content-Type: text/html; charset="US-ASCII" Content-Transfer-Encoding: quoted-printable

Hi experts,

I am working on the VBM analysis. For the data is = from multiple studies, some is sagitial and some is axial. After = “tbss_1” and check the slicedir, the data showed inconsistent orientation. Who = can help me and tell me how to change the = orientation?

Thanks a lot!

 

Xiaochu Zhang Ph.D

Visiting Research Fellow

NIH/NIDA-IRP

5500 Nathan Shock Drive

Baltimore<= font size=3D2 color=3Dnavy> = MD 21224

 

Tel: (410) - 550 - 1440 ext. 434

 

------_=_NextPart_001_01C80E01.41217E8D-- ========================================================================= Date: Sun, 14 Oct 2007 03:58:15 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: =?iso-8859-1?q?Gwena=EBlle=20DOUAUD?= <[log in to unmask]> Subject: Re: hi and for VBM problem. In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Hi Xiaochu, unfortunately, unless your lab has got a SGE system, that will not work in parallel on your server. Either you go for serial, then you will have to wait for a few days or alternatively, if you're familiar with scripting, you could also split your data in 16 bits and look in the fslvbm3 script what you should do to make it work. I would definitely recommend the first option! Cheers, Gwen --- "Zhang, Xiaochu (NIH/NIDA) [F]" <[log in to unmask]> a =E9crit : > Thank you very much! In our lab, there is a server > with 16 CPUs (the brand is HP). I always open a > terminal in my computer and do some analysis. > However, my computer was rebooted suddenly for some > reason. Therefore, I have to re-open a new terminal > and re-type the "fslvbm_3" again. Is your means that > the "fslvbm_3" was restart? And the time was before > reboot is no useful any more? > Thanks, again! > xiaochu > -----Original Message----- > From: Gwena=C3=ABlle DOUAUD > [mailto:[log in to unmask]]=20 > Sent: 2007=E5=B9=B410=E6=9C=8811=E6=97=A5 12:30 > To: [log in to unmask] > Subject: [FSL] RE : Re: [FSL] RE : Re: [FSL] RE : > [FSL] hi and for VBM problem. >=20 > If your lab has indeed a cluster (see your IT guys), > you could always restart it, it would only overwrite > what has been already created. I guess you can stop > it > by a simple ctrl c in the shell where you've typed > the > fslvbm3 command. But I would suggest you wait a bit, > as it should not be too long to give you the results > now.=20 >=20 > Gwenaelle >=20 > --- "Zhang, Xiaochu (NIH/NIDA) [F]" > <[log in to unmask]> a =C3=A9crit : >=20 > > Thank you very much!=20 > > So my question is whether I am able to interrupt > the > > program when it is running (For it took a long > time, > > sometimes it can not be avoid. ) and restart it > > again. > > Does it influence the final result? > > Thanks, again! > >=20 > > -----Original Message----- > > From: Gwena=C3=83=C2=ABlle DOUAUD > > [mailto:[log in to unmask]]=20 > > Sent: 2007=C3=A5=C2=B9=C2=B410=C3=A6=C5=93=CB=8611=C3=A6=E2=80=94=C2=A5= 10:36 > > To: [log in to unmask] > > Subject: [FSL] RE : Re: [FSL] RE : [FSL] hi and > for > > VBM problem. > >=20 > > Hi, > >=20 > > sorry, I should have been more specific... It can > be > > run in parallel if your lab has got a SGE system > or > > something like that (sorry I'm not very good at > > these > > things) and then you send your job to a bunch of > > computers involved in the cluster. It should not > > take > > much longer anyway.=20 > > Hopefully, we should get a better and much quicker > > non-linear registration (FNIRT) soonish and that > > will > > be included in the next FSL4.1 patch. > >=20 > > Cheers, > > Gwenaelle > >=20 > > =20 > > --- "Zhang, Xiaochu (NIH/NIDA) [F]" > > <[log in to unmask]> a =C3=83=C2=A9crit : > >=20 > > > Thank you very much for speedy response! > > > I think the first question has been resolved. > For > > 2 > > > hours per subject, it will take about 100 hours > > > (about 4 days) to finish the whole 48 subjects. > I > > > need wait two days more. > > > For you said it can be run in parallel and the > > > computer in our lab have multiple CPUs. So, can > I > > > open multiple terminals and run the script in > each > > > one to shorten the whole time? > > > Thanks again! > > >=20 > > > -----Original Message----- > > > From: Gwena=C3=83=C6=92=C3=82=C2=ABlle DOUAUD > > > [mailto:[log in to unmask]]=20 > > > Sent: > 2007=C3=83=C2=A5=C3=82=C2=B9=C3=82=C2=B410=C3=83=C2=A6=C3=85=E2=80=9C=C3= =8B=E2=80 11=C3=83=C2=A6=C3=A2=E2=82=AC=E2=80=9D=C3=82=C2=A5 > 9:39 > > > To: [log in to unmask] > > > Subject: [FSL] RE : [FSL] hi and for VBM > problem. > > >=20 > > > Hi, > > >=20 > > > I'm not sure about your first question. The > > > non-linear > > > registration can take a while, let's say at > least > > 2 > > > hours per subject.=20 > > > We have done the things so that you can run the > > jobs > > > in parallel if your lab has got a cluster to do > > so. > > > If > > > this is not the case, yes, it can be long, and > > much > > > longer than fslvbm2 for which you have hopefully > > > chosen the recommended affine option with -a. > > > Just check that your 4D image GM_mod_merg.nii.gz > > in > > > your stats directory has been created by the > most > > > recent fslvbm3 you ran, and if so, just abort > the > > > first fslvbm3 command.=20 > > >=20 > > > Hope this helps, > > > Gwenaelle > > >=20 > > >=20 > > > --- "Zhang, Xiaochu (NIH/NIDA) [F]" > > > <[log in to unmask]> a =C3=83=C6=92=C3=82=C2=A9crit : > > >=20 > > > > Hi, > > > >=20 > > > > I'm working on the VBM data. My sample > included > > 48 > > > > subjects (24 patients > > > > and 24 controls). I have two questions about > the > > > > "fslvbm_3_proc". > > > >=20 > > > > The first one, I ran it 2 days ago. However, > up > > to > > > > now, it have NOT > > > > finished. I checked the program is not died. > > > > However, when I run > > > > "fslvbm_1" and "fslvbm_2", it took no more > than > > 4 > > > > hours. Is it normal > > > > phenomenon? > > > >=20 > > > > The second one, about 1 day ago, my computer > was > > > > rebooted by power > > > > problem. After reboot, I re-run the > "fslvbm_3". > > It > > > > looks like OK. Need I > > > > to deleted all files and restart from > > "fslvbm_1"?=20 > > >=20 > > > >=20 > > > > Thank you very much! > > > >=20 > > > > Xiaochu Zhang Ph.D > > > >=20 > > > > Visiting Research Fellow > > > >=20 > > > > NIH/NIDA-IRP > > > >=20 > > > > 5500 Nathan Shock Drive > > > >=20 > > > > Baltimore MD 21224 > > > >=20 > > > > =20 > > > >=20 > > > > Tel: (410) - 550 - 1440 ext. 434 > > > >=20 > > > > =20 > > > >=20 > > > >=20 > > >=20 > > >=20 > > >=20 > > > =20 > > > > > > _________________________________________________________________________= ____ > > >=20 > > > Ne gardez plus qu'une seule adresse mail ! > Copiez > > > vos mails vers Yahoo! Mail=20 > > >=20 > >=20 > >=20 > >=20 > > =20 > > > _________________________________________________________________________= ____ > >=20 > > Ne gardez plus qu'une seule adresse mail ! Copiez > > vos mails vers Yahoo! Mail=20 > >=20 >=20 >=20 >=20 > =20 > _________________________________________________________________________= ____ >=20 > Ne gardez plus qu'une seule adresse mail ! Copiez > vos mails vers Yahoo! Mail=20 >=20 ___________________________________________________________________= __________=20 Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! M= ail=20 ========================================================================= Date: Sun, 14 Oct 2007 04:05:01 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: =?iso-8859-1?q?Gwena=EBlle=20DOUAUD?= <[log in to unmask]> Subject: RE : [FSL] The orientation problem in VBM analysis In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Hi again Xiaochu, for this orientational problem, please look at http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=3Dind0709&L=3DFSL&P=3DR4424= 8&I=3D-3 on the same subject... Cheers, Gwena=EBlle --- "Zhang, Xiaochu (NIH/NIDA) [F]" <[log in to unmask]> a =E9crit : > Hi experts, >=20 > I am working on the VBM analysis. For the data is > from multiple studies, > some is sagitial and some is axial. After "tbss_1" > and check the > slicedir, the data showed inconsistent orientation. > Who can help me and > tell me how to change the orientation? >=20 > Thanks a lot! >=20 > =20 >=20 > Xiaochu Zhang Ph.D >=20 > Visiting Research Fellow >=20 > NIH/NIDA-IRP >=20 > 5500 Nathan Shock Drive >=20 > Baltimore MD 21224 >=20 > =20 >=20 > Tel: (410) - 550 - 1440 ext. 434 >=20 > =20 >=20 >=20 ___________________________________________________________________= __________=20 Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! M= ail=20 ========================================================================= Date: Sun, 14 Oct 2007 11:35:34 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: Using SIENA in monkey MRIs In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, Is the movement between the two timepoints after registration due to atrophy effects of interest, or "real" misregistration? If you would like us to have a quick look, please upload the files in a single compressed tarfile to http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi and then email me the upload ID. Cheers. On 11 Oct 2007, at 03:15, Buyean Lee wrote: > Hi Steve, > > Finally, I ran Siena with the monkey MRIs (it is just one subject); > the siena_report.html looks much better in terms of brain > extraction, coregistation, etc. > > This is what I did. > > 1. removed all tissues outside the skull and below the cerebellum. > 2. modified the siena.sh in order to use separate BET option for > two scans. > > Brain extraction is much better, but coregistration by FLIRT is not > satisfactory. > I can still see some movement in the overlap of the two coregistred > MRIs. > > Is there any way for me to send these two MRIs to you so that you > can take a look at them? > > I just don't know what else I can do to improve the BET and > coregistration further. > > Thank you, > > Buyean > > > > -----Original Message----- > From: Steve Smith <[log in to unmask]> > To: [log in to unmask] > Sent: Thu, 27 Sep 2007 1:20 am > Subject: Re: [FSL] Using SIENA in monkey MRIs > > HI - you would also want to include a rasonable amount of the outer > skull boundary that's near the brain boundary, as this gets used in > SIENA. > > To choose a FOV in FSLView and then reduce the image by this: > Click around and find the voxel coordinates (voxel not mm) that > bound the region you want to keep. > Find the xmin and xmax values, and calculate the xsize = xmax - xmin > Same for y and z. > > Then run > > fslroi originalheadimage reducedimage xmin xsize ymin ysize zmin zsize > > (do this for both timepoints separately) > > and feed the reducedimages into SIENA. > > > Cheers. > > On 26 Sep 2007, at 23:52, Buyean Lee wrote: > > > Hi Steve, > > > > "Alternatively, you could work out how to reduce the field-of- > view > to just include the brain of each time point before feeding > the > output of that into SIENA. You would use FSLView on each > image to > work out the field-of-view and then use fslroi to reduce > the FOV of > the image." > > > > I created a whole brain mask for the animal brain using FSLView. > > > > But, it is not easy to understand how to use fslroi. > > > > fslroi ... > > > > I just don't know how to determine these parameters and when I am > > supposed to use the mask file in fslroi. > > > > Would you kindly let me know how to reduce the field-of-view to > > include the brain? > > > > > > Thank you, > > > > Buyean > > > ---------------------------------------------------------------------- > > ----- > > Check Out the new free AIM(R) Mail -- Unlimited storage and > > industry-leading spam and email virus protection. > > ---------------------------------------------------------------------- > ----- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > ---------------------------------------------------------------------- > ----- > Check Out the new free AIM(R) Mail -- Unlimited storage and > industry-leading spam and email virus protection. ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Sun, 14 Oct 2007 11:57:41 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: Problems with siena In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, sure, we can take a quick look: Please upload the files in a single compressed tarfile to http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi and then email me the upload ID. Cheers. On 11 Oct 2007, at 12:45, Shireen Kwiatkowska-Naqvi wrote: > Dear Mr Smith, > > I'm a newbie user of fsl, and I'm encountering some problems with > analysing data with siena: I've > tried out all the possible options found in the archives and > manual, but did not come to any good > results (the brain extraction does not work well) > > I'm using Siena to interpret pre- and post training T1 data, using > the -S command cannot get rid of > the eye region, combining different parameters does not help either > > Would it be possible to send you some sample data, so you could > take a look at it? > > With regards, > > Shireen Kwiatkowska-Naqvi > MA Student > Max-Planck-Institute for Development > Berlin ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Sun, 14 Oct 2007 14:09:23 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: volume of GM/WM/CSF from ROI In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, On 12 Oct 2007, at 16:47, Tamara Cristescu wrote: > Hi, > I would like to extract the amount of WM, GM and CSF from a > hippocampal ROI. > I have a T2 rat brain acquired at 0.160 x 0.160 x 0.320 (TE 55). I > multiply > the voxel size by 6 to get to 0.960 x 0.960 x 1.920 (otherwise I > can't do > anything with the data in FSL because voxels are too small). > I follow the standard analysis protocol: Bet, estimate bias field + > bias > field correction, and FAST on corrected volume. After FAST I get > rest1 I get > the pve files for each tissue (prest1_pve_1=GM, rest1_pve_2=WM and > rest1_pve_0=CSF) and the seg files (rest1_seg_1, rest1_seg_2 and > rest1_seg_0) and the the whole brain segmented image rest1_seg. > (rest1 is > the name of the volulme). I create a hippocampus mask called hippo- > mask > (intensity 1) and use fslstats to extract the info for GM: > fslstats rest1_pve_1 -l 1 -u 1 -k hippo-mask -m > 0.913644 > fslstats rest1_pve_1 -l 1 -u 1 -k hippo-mask -v > 6873 12161.580241 > I do the same for WM rest1_pve_2 (l=2 and u=2) and for CSF > rest1_pve_0 (l=0, > u=0). > I multiply the mean value (0.913644) with the 6873 to get the > volume in > voxels - 6,279 - or multiply the mean value 6873 with 12161.580241 > to get > the volume in cubic mm - 11,103. The values seem a little bit > unusual to me > for a small rat hippocampus. Really? Even if you divide by (6*6*6)? Apart from that it looks like your methodology is fine, as long as the segmentations look good by eye. Cheers. > I would like to extract WM, GM and CSF from the hipopcampus and > whole brain > volume to put them into a volumetric analysis. > I would be greateful for any suggestion on how to deal with this > data or any > alternatives for the volumetric measurement. > Tamara ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Sun, 14 Oct 2007 14:11:24 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: DTI processing In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, On 12 Oct 2007, at 18:06, Markus Gschwind wrote: > Hi FSLers! > > I am struggling with the DTI Mode of FSL: > > 1. > I can open FA data in FSLveiw 2D nicely, but I cannot open any FA- > data (produced in FSL3 on Cygwin) in FSLview 3D. That means: It > atually works but I just don't see anything! Is that normal? Have you tried setting the FSLView display range to 0:1? > 2. > I installed FSL 4 in a vmplayer. Atotest with the testdata worked > fine. > After Eddy corrent-correction (30 volumes of 30 have been > processed. "Done!") it tells me ERROR: > window name "prompt" already exists in parent > window name "prompt" already exists in parent > while executing > "toplevel .prompt -borderwidth 5" > (procedure "MxPause" line 4) > invoked from within > "MxPause " Done! " " > (procedure "fdt:apply" line 313) > invoked from within > "fdt:apply .fdt keep" > invoked from within > ".fdt.apply invoke" > ("uplevel" body line 1) > invoked from within > "uplevel #0 [list $w invoke]" > (procedure "tk::ButtonUp" line 22) > invoked from within > "tk::ButtonUp .fdt.apply" > (command bound to event) > and: Errors: terminate called after throwing an instance on > 'Newmat::InompatibleDimensionsException'. I think we'd need to see more info about what aspects of the analysis worked and what didn't? Possibly you had one of the popup windows still up when it was supposed to have been closed? Sounds like the analysis did run though? Cheers. > 3. Trying Bedpostx on another data I got THE SAME error message. > > What is wrong? How can I fix that? > > Thank you very much! > Cheers, > Markus > > -- > > Markus Gschwind, M.D. > Laboratory for Neurology and Imaging of Cognition > Dept of Neurosciences > University Medical Center (CMU) > 1 Michel-Servet - 1211 GENEVA - CH > > Tel 0041 (0) 22 379 5324 > Fax 0041 (0) 22 379 5402 > email: [log in to unmask] > http://labnic.unige.ch ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Sun, 14 Oct 2007 14:13:10 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: installing problem of fsl 4.0 In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - the Apple binaries are compiled for Tiger (10.4) - almost no-one uses the previous OS now as it's so old. Hopefully though you should be able to compile from the source FSL distribution. Sorry not to have an easier fix to suggest! Cheers, Steve. On 12 Oct 2007, at 22:39, Junghee Lee wrote: > Hello Everyone, > > I've tried to install FSL 4.0 on Apple G5 running OS X 10.3.8. I > used the installer script and didn't have any problem until I tried > to launch fsl gui. > > When I tried to use GUI, I got the following error message. I even > tried to run FSL GUI straight from X 11 and got the same message. > > JQLab-G530:~/Desktop/Download jedimaster$ dyld: /usr/local/fsl/ > extras/bin/fslwish8.4 Undefined symbols: > /usr/local/fsl/extras/lib/libtcl8.4.dylib undefined reference to > _OSSpinLockLock expected to be defined in /usr/lib/libSystem.B.dylib > > > I also copied the installation process below. Any feedback would > be greatly appreciated. > > Junghee > > > JQLab-G530:~/Desktop/Download jedimaster$ sudo sh fsl_installer.sh > Password: > Looking for FSL tarball in the current directory... > *************************************************************** > No FSL tarball specified, assuming you want me to install > /Users/jedimaster/Desktop/Download/fsl-4.0.1-macosx.tar > from the current directory. > *************************************************************** > > FSL instal script > ================= > > Where would you like to install FSL to? [/usr/local] > '/usr/local/fsl' exists - would you like to remove it? [yes] > Deleting '/usr/local/fsl'... > Installing FSL from /Users/jedimaster/Desktop/Download/fsl-4.0.1- > macosx.tar into '/usr/local'... > > Untarring fsl-4.0.1-macosx.tar > Would you like to install FSLView.app into /Applications (requires > Administrator priviledges)? [yes] > Moving old FSLView into the Trash > Installing FSLView into /Applications > Patching environment > '.bash_profile' already contains a FSLDIR definition which is > correctly configured, not changing. > Setting up Apple terminal... > DISPLAY is already being configured in your '/Users/ > jedimaster/.bash_profile' - not changing > JQLab-G530:~/Desktop/Download jedimaster$ $FSLDIR > -bash: /usr/local/fsl: is a directory > JQLab-G530:~/Desktop/Download jedimaster$ flirt -version > FLIRT version 5.4.2 > JQLab-G530:~/Desktop/Download jedimaster$ fsl & > [1] 742 > JQLab-G530:~/Desktop/Download jedimaster$ dyld: /usr/local/fsl/ > extras/bin/fslwish8.4 Undefined symbols: > /usr/local/fsl/extras/lib/libtcl8.4.dylib undefined reference to > _OSSpinLockLock expected to be defined in /usr/lib/libSystem.B.dylib > > [1]+ Trace/BPT trap fsl > > > On Oct 1, 2007, at 4:35 AM, Matthew Webster wrote: > >> Hi, >> To make X11 terminal run fsl you need to add -ls to the >> xterm command than X11 runs ( go to X11-Applications-Customize >> Menu and change xterm ) to make it read your .profile when >> starting a shell.. >> >> Matthew >>> Hello, >>> >>> I've just installed FSL on my new Intel Mac running OS X Tiger. >>> I used the >>> installer script and installed FSL into the /Applications >>> directory. When I >>> run the tests using the $FSLDIR command, etc. in terminal all is >>> fine, >>> however, when I go to X11 and type fsl I get the error that >>> "bash: fsl: >>> command not found". I can launch fsl from within terminal by >>> typing fsl just >>> not from the X11 terminal. >>> >>> Also when I open terminal I used to get the following error: >>> /Applications/fsl/etc/fslconf/fslmachtype.sh: line 1: gcc: >>> command not >>> found. I found an e-mail in the archive which advised editing the >>> fslmachtype.sh file which I have done. I now get a new error >>> that says: >>> /Applications/fsl/etc/fslconf/fslmachtype.sh: line 27: [0: >>> command not found. >>> >>> I am not sure if the above two problems are connected or what I >>> should do >>> about either of them. Thank in advance for the help. >>> >>> Liz >> >> > ------------------------------------ > Junghee Lee > Post-doctoral fellow > Semel Institute of Neuroscience and Human Behavior, UCLA > > UCLA/VA Greater Los Angeles Healthcare System > 11301 Wilshire Blvd. > MIRECC, Bldg. 210 (MC 210A) > Los Angeles, CA 90073 > > 310-478-3711 ext.42986 > http://greenlab.npih.ucla.edu ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Sun, 14 Oct 2007 14:18:47 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: fslvm32 Comments: cc: Duncan Mortimer <[log in to unmask]> In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit I think that;s right, yes - Duncan can correct me if I'm wrong - I doubt this will give you 64-bit functionality. If you are needing to do large analyses we would strongly recommend against using Windows anyway - I would recommend properly native Linux or Apple. Cheers, Steve. On 13 Oct 2007, at 17:43, Emily RUBIN-FERREIRA wrote: > Hi, > I would like to install FSL 4.0 on a Windows XP computer 64-bit, using > vmware and fslvm32 as instructed. > > I was reading in the FAQ the following: that on 32-bit computers, > there > is a limit on the size that a single program can be regardless of the > RAM/swap you have, and that the limit is 2 GB. If fslvm is 32bit, does > this mean that if I install using fslvm, that my computer becomes > like a > 32-bit computer? > > Thanks, > emily ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Sun, 14 Oct 2007 15:21:29 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Jiri Keller <[log in to unmask]> Subject: Re: dtifit throwing instance In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline 2007/10/14, Jiri Keller <[log in to unmask]>: > Dear Tim, > > > can you run bedpostx_datacheck on the directory? > Thank you, I was not aware of this tool, it revealed bvecs/bvals > discrepancy, tiny check and editing and it is working ! > > with thanks > George > ========================================================================= Date: Sun, 14 Oct 2007 11:18:28 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Emily RUBIN-FERREIRA <[log in to unmask]> Subject: Re: fslvm32 In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Great thanks very much for the response. I'm sure we'll switch to Linux but in a last ditch effort to continue using windows, are there are any plans to come out with fslvm64? Thanks, emily On Sun, 14 Oct 2007, Steve Smith wrote: > I think that;s right, yes - Duncan can correct me if I'm wrong - I > doubt this will give you 64-bit functionality. If you are needing to > do large analyses we would strongly recommend against using Windows > anyway - I would recommend properly native Linux or Apple. > > Cheers, Steve. > > > On 13 Oct 2007, at 17:43, Emily RUBIN-FERREIRA wrote: > > > Hi, > > I would like to install FSL 4.0 on a Windows XP computer 64-bit, using > > vmware and fslvm32 as instructed. > > > > I was reading in the FAQ the following: that on 32-bit computers, > > there > > is a limit on the size that a single program can be regardless of the > > RAM/swap you have, and that the limit is 2 GB. If fslvm is 32bit, does > > this mean that if I install using fslvm, that my computer becomes > > like a > > 32-bit computer? > > > > Thanks, > > emily > > > ------------------------------------------------------------------------ > --- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > ------------------------------------------------------------------------ > --- > ========================================================================= Date: Sun, 14 Oct 2007 21:05:23 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Ping-Hong Yeh <[log in to unmask]> Subject: mm to vox coordinate Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Hi,=20 I'd like to know the antomate way of converting the coordinates in mm to = vox so I could extract the ROI using fslroi. I knew I can just open the fslvi= ew and get the vox, but would perfer not to do so. I have used the SPM scrip= ts (i.e. mm =3D [-22, 50, 50]; % example co-ordinates. V =3D spm_vol(spm_get(1,'*.img')); M =3D inv(V.mat); vox =3D (M(1:3,1:3)*mm' + repmat(M(1:3,4),1,size(mm,1)))', ) but it did not get the right vox coordinate, not knowing why? Thanks.=20 Ping ========================================================================= Date: Sun, 14 Oct 2007 16:16:58 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matt Glasser <[log in to unmask]> Organization: ma-tea.com Subject: Re: mm to vox coordinate In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Ping, I just wrote a script to convert MNI coordinates in mm to voxel coordinates of the FSL standard brain. The following shell script code takes the 1st 2nd and 3rd commandline arguments and converts them to voxel coordinates in variables x y z: x=`echo "$1 * -1 + 90" | bc` y=`echo "$2 * 1 + 126" | bc` z=`echo "$3 * 1 + 72" | bc` If you just want to know what the coordinates are, you could add the line: echo $x $y $z and give the script a name. Then you would type ./script.sh where the last three values are the MNI coordinates in mm. I use the script to draw little spheres at particular coordinates when plotting functional activations in the literature. Peace, Matt. -----Original Message----- From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Ping-Hong Yeh Sent: Sunday, October 14, 2007 4:05 PM To: [log in to unmask] Subject: [FSL] mm to vox coordinate Hi, I'd like to know the antomate way of converting the coordinates in mm to vox so I could extract the ROI using fslroi. I knew I can just open the fslview and get the vox, but would perfer not to do so. I have used the SPM scripts (i.e. mm = [-22, 50, 50]; % example co-ordinates. V = spm_vol(spm_get(1,'*.img')); M = inv(V.mat); vox = (M(1:3,1:3)*mm' + repmat(M(1:3,4),1,size(mm,1)))', ) but it did not get the right vox coordinate, not knowing why? Thanks. Ping ========================================================================= Date: Sun, 14 Oct 2007 21:52:49 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mark Jenkinson <[log in to unmask]> Subject: Re: mm to vox coordinate In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Dear Ping, You can use the command: std2imgcoord to convert between coordinates (also img2imgcoord and img2stdcoord). These are easy to incorporate within scripts. All the best, Mark On 14 Oct 2007, at 21:05, Ping-Hong Yeh wrote: > Hi, > > I'd like to know the antomate way of converting the coordinates in > mm to vox > so I could extract the ROI using fslroi. I knew I can just open the > fslview > and get the vox, but would perfer not to do so. I have used the SPM > scripts > (i.e. mm = [-22, 50, 50]; % example co-ordinates. > V = spm_vol(spm_get(1,'*.img')); > M = inv(V.mat); > vox = (M(1:3,1:3)*mm' + repmat(M(1:3,4),1,size(mm,1)))', ) > > but it did not get the right vox coordinate, not knowing why? > > Thanks. > > Ping ========================================================================= Date: Sun, 14 Oct 2007 17:03:16 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matt Glasser <[log in to unmask]> Organization: ma-tea.com Subject: Re: mm to vox coordinate In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Would be nice if that were documented somewhere. :) -----Original Message----- From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Mark Jenkinson Sent: Sunday, October 14, 2007 4:53 PM To: [log in to unmask] Subject: Re: [FSL] mm to vox coordinate Dear Ping, You can use the command: std2imgcoord to convert between coordinates (also img2imgcoord and img2stdcoord). These are easy to incorporate within scripts. All the best, Mark On 14 Oct 2007, at 21:05, Ping-Hong Yeh wrote: > Hi, > > I'd like to know the antomate way of converting the coordinates in > mm to vox > so I could extract the ROI using fslroi. I knew I can just open the > fslview > and get the vox, but would perfer not to do so. I have used the SPM > scripts > (i.e. mm = [-22, 50, 50]; % example co-ordinates. > V = spm_vol(spm_get(1,'*.img')); > M = inv(V.mat); > vox = (M(1:3,1:3)*mm' + repmat(M(1:3,4),1,size(mm,1)))', ) > > but it did not get the right vox coordinate, not knowing why? > > Thanks. > > Ping ========================================================================= Date: Sun, 14 Oct 2007 22:13:51 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mark Jenkinson <[log in to unmask]> Subject: Re: mm to vox coordinate In-Reply-To: [log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Dear Matt, It has been - under misc flirt utilities linked near the top of the main flirt page (as flirt is the main tool dealing with registration and coordinates). See: http://www.fmrib.ox.ac.uk/fsl/flirt/overview.html#misc All the best, Mark On 14 Oct 2007, at 22:03, Matt Glasser wrote: > Would be nice if that were documented somewhere. :) > > -----Original Message----- > From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On > Behalf > Of Mark Jenkinson > Sent: Sunday, October 14, 2007 4:53 PM > To: [log in to unmask] > Subject: Re: [FSL] mm to vox coordinate > > Dear Ping, > > You can use the command: > std2imgcoord > to convert between coordinates (also img2imgcoord and img2stdcoord). > These are easy to incorporate within scripts. > > All the best, > Mark > > > On 14 Oct 2007, at 21:05, Ping-Hong Yeh wrote: > >> Hi, >> >> I'd like to know the antomate way of converting the coordinates in >> mm to vox >> so I could extract the ROI using fslroi. I knew I can just open the >> fslview >> and get the vox, but would perfer not to do so. I have used the SPM >> scripts >> (i.e. mm = [-22, 50, 50]; % example co-ordinates. >> V = spm_vol(spm_get(1,'*.img')); >> M = inv(V.mat); >> vox = (M(1:3,1:3)*mm' + repmat(M(1:3,4),1,size(mm,1)))', ) >> >> but it did not get the right vox coordinate, not knowing why? >> >> Thanks. >> >> Ping ========================================================================= Date: Mon, 15 Oct 2007 02:25:21 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Ping-Hong Yeh <[log in to unmask]> Subject: Re: mm to vox coordinate Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Hi Matt and Mark,=20 Thanks for the reply.=20 I only need to know the correpsonding vox for the center of a ROI so I ca= n use the fslroi to get the tissue imformation after I ran the fast. Since = the MPRAGE (256*256*154) were in the native space and I will not transform th= em to the MNI standard space, I don't think the std2imgcoord or img2imgcoord= are what I need.=20 I am wondering how fslview get the coordianates in mm of (0,0,0), which varies between the image even though they were inthe same resolution?=20 I can transform the mm to vox once I knew these x y z .=20 Regards,=20 Ping =20=20 On Sun, 14 Oct 2007 22:13:51 +0100, Mark Jenkinson <[log in to unmask]> = wrote: >Dear Matt, > >It has been - under misc flirt utilities linked near the top of the main= >flirt page (as flirt is the main tool dealing with registration and >coordinates). >See: > http://www.fmrib.ox.ac.uk/fsl/flirt/overview.html#misc > >All the best, >=09Mark > > > >On 14 Oct 2007, at 22:03, Matt Glasser wrote: > >> Would be nice if that were documented somewhere. :) >> >> -----Original Message----- >> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On >> Behalf >> Of Mark Jenkinson >> Sent: Sunday, October 14, 2007 4:53 PM >> To: [log in to unmask] >> Subject: Re: [FSL] mm to vox coordinate >> >> Dear Ping, >> >> You can use the command: >> =09std2imgcoord >> to convert between coordinates (also img2imgcoord and img2stdcoord). >> These are easy to incorporate within scripts. >> >> All the best, >> =09Mark >> >> >> On 14 Oct 2007, at 21:05, Ping-Hong Yeh wrote: >> >>> Hi, >>> >>> I'd like to know the antomate way of converting the coordinates in >>> mm to vox >>> so I could extract the ROI using fslroi. I knew I can just open the >>> fslview >>> and get the vox, but would perfer not to do so. I have used the SPM >>> scripts >>> (i.e. mm =3D [-22, 50, 50]; % example co-ordinates. >>> V =3D spm_vol(spm_get(1,'*.img')); >>> M =3D inv(V.mat); >>> vox =3D (M(1:3,1:3)*mm' + repmat(M(1:3,4),1,size(mm,1)))', ) >>> >>> but it did not get the right vox coordinate, not knowing why? >>> >>> Thanks. >>> >>> Ping ========================================================================= Date: Mon, 15 Oct 2007 02:36:41 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Marc Dubin <[log in to unmask]> Subject: mean_FA_skeleton and brain misalignment Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain Hello FSL Users, Using the TBSS tools I have been carrying out the tbss_2_reg step with th= e -T option (FMRIB58_FA). I=20 then go through all of the remaining steps of TBSS and then when I visual= ise with: fslview MNI152 mean_FA_skeleton -l Green -b 2000,8000 tbss_tstat1 -l Red-= Yellow -b 3,6=20 tbss_tstat2 -l Blue-Lightblue -b 3,6 the mean_FA_skeleton significantly overshoots the brain, terminating in o= r even beyond the skull=20 (the overall extent of the skeleton is significantly larger that the brai= n). Any ideas about what might be going on here would be greatly appreciated!= Thanks, Marc ========================================================================= Date: Sun, 14 Oct 2007 22:01:41 -0500 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Catherine Warrier <[log in to unmask]> Subject: simple fslmaths -roi question Content-Type: text/plain Content-Disposition: inline Content-Transfer-Encoding: binary MIME-Version: 1.0 Dear list, I am brand new to FSL and am having trouble using the fslmaths -roi function. I have a number of volumes with two regions of interest each, one on the right and one on the left. I need to extract data from the left and right separately and would like to avoid manually erasing each one separately. I understand that this function sets everything outside the designated region to 0, so I want to select one hemisphere at a time. I thought I understood this but I can only seem to generate empty volumes. Here's the syntax: fslmaths input -roi output Here's an example of what I've tried: fslmaths input -roi -75 75 -40 160 -40 130 0 1 output specifying the left hemisphere and the entire y and z dimensions. I only have one volume, so tried timepoint 0 with 1 timepoint as well as 1 and 1. I also tried decreasing the extent of the y and z dimensions. Tried playing with the min and max of the viewer when viewing the resulting volume, but it is truly empty. Tried the function with and without extensions on the input/output files... What am I doing wrong? Any suggestions would be greatly appreciated. Thank you, Catherine ~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~- Catherine Warrier, PhD Research Assistant Professor, Knowles Fellow Auditory Neuroscience Laboratory Northwestern University ~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~- ========================================================================= Date: Mon, 15 Oct 2007 07:11:10 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: mm to vox coordinate In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - I'm confused - if you are using FSLView to find the co-ordinates in the first place, and then need to use voxel co-ordinates (as opposed to mm) then you can just read the voxel co-ords in FSLView directly (just to the left of the mm co-ords). Cheers. On 15 Oct 2007, at 02:25, Ping-Hong Yeh wrote: > Hi Matt and Mark, > > Thanks for the reply. > I only need to know the correpsonding vox for the center of a ROI > so I can > use the fslroi to get the tissue imformation after I ran the fast. > Since the > MPRAGE (256*256*154) were in the native space and I will not > transform them > to the MNI standard space, I don't think the std2imgcoord or > img2imgcoord > are what I need. > I am wondering how fslview get the coordianates in mm of (0,0,0), > which > varies between the image even though they were inthe same resolution? > I can transform the mm to vox once I knew these x y z . > Regards, > > Ping > > > > On Sun, 14 Oct 2007 22:13:51 +0100, Mark Jenkinson > <[log in to unmask]> wrote: > >> Dear Matt, >> >> It has been - under misc flirt utilities linked near the top of >> the main >> flirt page (as flirt is the main tool dealing with registration and >> coordinates). >> See: >> http://www.fmrib.ox.ac.uk/fsl/flirt/overview.html#misc >> >> All the best, >> Mark >> >> >> >> On 14 Oct 2007, at 22:03, Matt Glasser wrote: >> >>> Would be nice if that were documented somewhere. :) >>> >>> -----Original Message----- >>> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On >>> Behalf >>> Of Mark Jenkinson >>> Sent: Sunday, October 14, 2007 4:53 PM >>> To: [log in to unmask] >>> Subject: Re: [FSL] mm to vox coordinate >>> >>> Dear Ping, >>> >>> You can use the command: >>> std2imgcoord >>> to convert between coordinates (also img2imgcoord and img2stdcoord). >>> These are easy to incorporate within scripts. >>> >>> All the best, >>> Mark >>> >>> >>> On 14 Oct 2007, at 21:05, Ping-Hong Yeh wrote: >>> >>>> Hi, >>>> >>>> I'd like to know the antomate way of converting the coordinates in >>>> mm to vox >>>> so I could extract the ROI using fslroi. I knew I can just open the >>>> fslview >>>> and get the vox, but would perfer not to do so. I have used the SPM >>>> scripts >>>> (i.e. mm = [-22, 50, 50]; % example co-ordinates. >>>> V = spm_vol(spm_get(1,'*.img')); >>>> M = inv(V.mat); >>>> vox = (M(1:3,1:3)*mm' + repmat(M(1:3,4),1,size(mm,1)))', ) >>>> >>>> but it did not get the right vox coordinate, not knowing why? >>>> >>>> Thanks. >>>> >>>> Ping ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Mon, 15 Oct 2007 07:16:59 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: simple fslmaths -roi question In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - sorry, these dimensions are in voxels not mm - so if you're estimating the ROI from FSLView you need the leftmost co-ordinates (voxels) not the mm co-ords. The voxel co-ordinate frame starts at [0,0,0] in the corner of the image and never goes negative. Cheers. On 15 Oct 2007, at 04:01, Catherine Warrier wrote: > Dear list, > > I am brand new to FSL and am having trouble using the fslmaths -roi > function. I have a number of > volumes with two regions of interest each, one on the right and one > on the left. I need to extract > data from the left and right separately and would like to avoid > manually erasing each one > separately. I understand that this function sets everything > outside the designated region to 0, so > I want to select one hemisphere at a time. I thought I understood > this but I can only seem to > generate empty volumes. > > Here's the syntax: > fslmaths input -roi > output > > Here's an example of what I've tried: > fslmaths input -roi -75 75 -40 160 -40 130 0 1 output > > specifying the left hemisphere and the entire y and z dimensions. > I only have one volume, so tried > timepoint 0 with 1 timepoint as well as 1 and 1. I also tried > decreasing the extent of the y and z > dimensions. Tried playing with the min and max of the viewer when > viewing the resulting volume, > but it is truly empty. Tried the function with and without > extensions on the input/output files... > > What am I doing wrong? Any suggestions would be greatly appreciated. > > Thank you, > Catherine > > > ~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~- > Catherine Warrier, PhD > Research Assistant Professor, Knowles Fellow > Auditory Neuroscience Laboratory > Northwestern University > ~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~- ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Mon, 15 Oct 2007 07:19:51 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: mean_FA_skeleton and brain misalignment In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - I'm suprised that you get this with the -T option. If you load in all_FA first into FSLView, and then overlay the FMRIB58_FA (threshold this, change its colour to red-yellow, and make it a little transparent) and then move through the timepoints, do all subjects behave in a similar way to what you've described? We can have a look and advise on what's happening; Please upload the files in a single compressed tarfile to http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi and then email me the upload ID. Cheers. On 15 Oct 2007, at 02:36, Marc Dubin wrote: > Hello FSL Users, > > Using the TBSS tools I have been carrying out the tbss_2_reg step > with the -T option (FMRIB58_FA). I > then go through all of the remaining steps of TBSS and then when I > visualise with: > > fslview MNI152 mean_FA_skeleton -l Green -b 2000,8000 tbss_tstat1 - > l Red-Yellow -b 3,6 > tbss_tstat2 -l Blue-Lightblue -b 3,6 > > the mean_FA_skeleton significantly overshoots the brain, > terminating in or even beyond the skull > (the overall extent of the skeleton is significantly larger that > the brain). > > Any ideas about what might be going on here would be greatly > appreciated! > > Thanks, > Marc ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Mon, 15 Oct 2007 09:08:50 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Vincent Keereman <[log in to unmask]> Subject: FW: Tissue probability map for bone structures MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0040_01C80F0B.00631710" This is a multipart message in MIME format. ------=_NextPart_000_0040_01C80F0B.00631710 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Hello, I'm segmenting MR brain scans, but apart from WM, GM and CSF, I also need to locate the bone structures correctly. Does anyone know if there are any tissue probability maps/templates available for bone? BR, Vincent ------=_NextPart_000_0040_01C80F0B.00631710 Content-Type: text/html; charset="US-ASCII" Content-Transfer-Encoding: quoted-printable

Hello,

 

I’m segmenting MR brain = scans, but apart from WM, GM and CSF, I also need to locate the bone structures  = correctly. Does anyone know if there are any tissue probability maps/templates = available for bone?

 

BR,

 

Vincent

------=_NextPart_000_0040_01C80F0B.00631710-- ========================================================================= Date: Mon, 15 Oct 2007 09:29:48 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Mark Jenkinson <[log in to unmask]> Subject: Re: mm to vox coordinate In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, The (0,0,0) coordinate is encoded by the nifti files and most likely comes from the scanner coordinate system. In that case it encodes where the scanner believed the centre of the scanner's coordinates were within the image. Hence this may not be of interest to you. If, for any reason, you still do want to know where that is without using fslview, then you can use: std2imgcoord -img image_name -vox coord_file to extract corresponding voxel coordinates to the mm coordinates contained in the file coord_file (one set of 3 values per line) which will use the same calculations that FSLView does to convert between mm and vox. Note that this does not rely on the image being related to std space in this case - it will work for any image. Hope this is of some help. All the best, Mark On 15 Oct 2007, at 02:25, Ping-Hong Yeh wrote: > Hi Matt and Mark, > > Thanks for the reply. > I only need to know the correpsonding vox for the center of a ROI > so I can > use the fslroi to get the tissue imformation after I ran the fast. > Since the > MPRAGE (256*256*154) were in the native space and I will not > transform them > to the MNI standard space, I don't think the std2imgcoord or > img2imgcoord > are what I need. > I am wondering how fslview get the coordianates in mm of (0,0,0), > which > varies between the image even though they were inthe same resolution? > I can transform the mm to vox once I knew these x y z . > Regards, > > Ping > > > > On Sun, 14 Oct 2007 22:13:51 +0100, Mark Jenkinson > <[log in to unmask]> wrote: > >> Dear Matt, >> >> It has been - under misc flirt utilities linked near the top of >> the main >> flirt page (as flirt is the main tool dealing with registration and >> coordinates). >> See: >> http://www.fmrib.ox.ac.uk/fsl/flirt/overview.html#misc >> >> All the best, >> Mark >> >> >> >> On 14 Oct 2007, at 22:03, Matt Glasser wrote: >> >>> Would be nice if that were documented somewhere. :) >>> >>> -----Original Message----- >>> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On >>> Behalf >>> Of Mark Jenkinson >>> Sent: Sunday, October 14, 2007 4:53 PM >>> To: [log in to unmask] >>> Subject: Re: [FSL] mm to vox coordinate >>> >>> Dear Ping, >>> >>> You can use the command: >>> std2imgcoord >>> to convert between coordinates (also img2imgcoord and img2stdcoord). >>> These are easy to incorporate within scripts. >>> >>> All the best, >>> Mark >>> >>> >>> On 14 Oct 2007, at 21:05, Ping-Hong Yeh wrote: >>> >>>> Hi, >>>> >>>> I'd like to know the antomate way of converting the coordinates in >>>> mm to vox >>>> so I could extract the ROI using fslroi. I knew I can just open the >>>> fslview >>>> and get the vox, but would perfer not to do so. I have used the SPM >>>> scripts >>>> (i.e. mm = [-22, 50, 50]; % example co-ordinates. >>>> V = spm_vol(spm_get(1,'*.img')); >>>> M = inv(V.mat); >>>> vox = (M(1:3,1:3)*mm' + repmat(M(1:3,4),1,size(mm,1)))', ) >>>> >>>> but it did not get the right vox coordinate, not knowing why? >>>> >>>> Thanks. >>>> >>>> Ping ========================================================================= Date: Mon, 15 Oct 2007 09:31:15 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Alle Meije Wink <[log in to unmask]> Subject: Re: DICOM converter MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit For everything from thousands of single-slice DICOM to 4D images like Philips's PAR/REC which is sort-of DICOM, I use the dcm2nii command that comes with MriCRoN (http://www.sph.sc.edu/comd/rorden/mricron). As for the "argument list too long" problem, when I have a directory full of 2D dicom files that together constitute a 4D data set, calling "dcm2nii " automatically packs the data together in 1 file. Has never failed so far. It sets the q-form (which for fMRI is not really useful but it's easy to set to 0) and other NifTI fields. The slice_order field must be set manually (for programs that can use it). hth Alle Meije ___________________________________________________________ Yahoo! Answers - Got a question? Someone out there knows the answer. Try it now. http://uk.answers.yahoo.com/ ========================================================================= Date: Mon, 15 Oct 2007 10:00:18 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: FW: Tissue probability map for bone structures In-Reply-To: <003f01c80efa$3cda4710$b68ed530$@[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=WINDOWS-1252; delsp=yes; format=flowed Content-Transfer-Encoding: quoted-printable Hi, This is generally hard as no readily-available MR sequences show up =20 bone as anything but dark; it's much easier with CT. However, if you =20 get good quality T1 _AND_ T2 you can use BET2 to get an estimation of =20= where the skull's internal and external surfaces are, which generally =20= works well for the majority of the head. See the BET manual page for =20= more information. Cheers, Steve. On 15 Oct 2007, at 08:08, Vincent Keereman wrote: > Hello, > > > > I=92m segmenting MR brain scans, but apart from WM, GM and CSF, I =20 > also need to locate the bone structures correctly. Does anyone =20 > know if there are any tissue probability maps/templates available =20 > for bone? > > > > BR, > > > > Vincent > > ------------------------------------------------------------------------=20= --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------=20= --- ========================================================================= Date: Mon, 15 Oct 2007 09:52:27 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Patrick Hales <[log in to unmask]> Subject: Problem with DTIFit Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain Hi This is the first time I have tried using FSL, and I am trying to get the= DTIFit program to work. I have=20 a series of diffusion weighted images, each of which are stored as indivi= dual fid files (created using=20 Varian's VnmrJ software). There are seven images in total, one with b=3D0= and 6 with diffusion=20 gradients applied in non co-linear directions. I have converted each of = these into Nifti format, and=20 stored them all in a directory. I used FSLview to create my binary mask, = and have created text files=20 with my gradient directions (all normalized to unit length), and my b-val= ues (in units of s/mm2). In=20 the DTIFit program, I have selected 'specify input files manually', and s= upplied it with the directory=20 containing my nifti files, and the filenames of my gradients and b-values= text files. When I press 'go', I get a number of error messages, such as: 'ERROR: nif= ti_image_read(....directory=20 name....)can't open header file', 'ERROR: nifti_image_open(...):bad heade= r info', 'Error: failed to open=20 file', etc. I get the same errors when I try it with analyze files. Would you be able to tell me what I'm doing wrong? Thanks very much Patrick Hales ========================================================================= Date: Mon, 15 Oct 2007 12:04:42 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Markus Gschwind <[log in to unmask]> Subject: Re: DTI processing In-Reply-To: <[log in to unmask]> MIME-version: 1.0 Content-transfer-encoding: 7BIT Content-disposition: inline Content-type: text/plain; charset=ISO-8859-1; DelSp=Yes; format=flowed Hi Steve! Thanks for the answers. Quoting Steve Smith <[log in to unmask]>: >> 1. >> I can open FA data in FSLveiw 2D nicely, but I cannot open any >> FA-data (produced in FSL3 on Cygwin) in FSLview 3D. That means: It >> atually works but I just don't see anything! Is that normal? > > Have you tried setting the FSLView display range to 0:1? Yes exactly. But I dont see anything. Acutally in the terminal it tells me: "ERROR: In /home/duncan/VTK/Graphics/vtkPolyDataNormals.cxx, line 94 vtkPolyDataNormals (0xa2fccb8): No data to generate normals for!" Is that a compatibility problem of FSL 3 and FSL 4? Thanks again, cheers, Markus >> -- >> >> Markus Gschwind, M.D. >> Laboratory for Neurology and Imaging of Cognition >> Dept of Neurosciences >> University Medical Center (CMU) >> 1 Michel-Servet - 1211 GENEVA - CH >> >> Tel 0041 (0) 22 379 5324 >> Fax 0041 (0) 22 379 5402 >> email: [log in to unmask] >> http://labnic.unige.ch > > > --------------------------------------------------------------------------- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > --------------------------------------------------------------------------- -- Markus Gschwind, M.D. Laboratory for Neurology and Imaging of Cognition Dept of Neurosciences University Medical Center (CMU) 1 Michel-Servet - 1211 GENEVA - CH Tel 0041 (0) 22 379 5324 Fax 0041 (0) 22 379 5402 email: [log in to unmask] http://labnic.unige.ch ========================================================================= Date: Mon, 15 Oct 2007 11:14:53 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Tim Behrens <[log in to unmask]> Subject: Re: Problem with DTIFit In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit You need to merge your data from different diffusion directions into a single 4D nifti file - you can use fslmerge to do this. T On 15 Oct 2007, at 09:52, Patrick Hales wrote: > Hi > > This is the first time I have tried using FSL, and I am trying to > get the DTIFit program to work. I have > a series of diffusion weighted images, each of which are stored as > individual fid files (created using > Varian's VnmrJ software). There are seven images in total, one with > b=0 and 6 with diffusion > gradients applied in non co-linear directions. I have converted > each of these into Nifti format, and > stored them all in a directory. I used FSLview to create my binary > mask, and have created text files > with my gradient directions (all normalized to unit length), and my > b-values (in units of s/mm2). In > the DTIFit program, I have selected 'specify input files manually', > and supplied it with the directory > containing my nifti files, and the filenames of my gradients and b- > values text files. > > When I press 'go', I get a number of error messages, such as: > 'ERROR: nifti_image_read(....directory > name....)can't open header file', 'ERROR: nifti_image_open(...):bad > header info', 'Error: failed to open > file', etc. I get the same errors when I try it with analyze files. > > Would you be able to tell me what I'm doing wrong? Thanks very much > > Patrick Hales ========================================================================= Date: Mon, 15 Oct 2007 11:18:16 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Saad Jbabdi <[log in to unmask]> Subject: Re: Problem with DTIFit In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi On 15 Oct 2007, at 09:52, Patrick Hales wrote: > Hi > > This is the first time I have tried using FSL, and I am trying to > get the DTIFit program to work. I have > a series of diffusion weighted images, each of which are stored as > individual fid files (created using > Varian's VnmrJ software). There are seven images in total, one with > b=0 and 6 with diffusion > gradients applied in non co-linear directions. I have converted > each of these into Nifti format, and > stored them all in a directory. I used FSLview to create my binary > mask, and have created text files > with my gradient directions (all normalized to unit length), and my > b-values (in units of s/mm2). In > the DTIFit program, I have selected 'specify input files manually', > and supplied it with the directory > containing my nifti files, and the filenames of my gradients and b- > values text files. Sorry it is not clear here whether you selected each file manually (separately), or just provided the name of the directory containing the data files, in which case the files have to have special names (data,bvecs,bvals,nodif_brain_mask) ? Saad. ========================================================================= Date: Mon, 15 Oct 2007 11:36:25 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Patrick Hales <[log in to unmask]> Subject: Re: Problem with DTIFit In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi Thanks for that. The command 'fslmerge' isn't recognized for some reason, however, I have been able to make my 4D data file using 'avwmerge' instead. Unfortunately, when I try to process the data in DTIFit, I now get the error message: "Error: child killed: bus error". Would you be able to tell me how to fix this? Thanks very much for your help Patrick On 15 Oct 2007, at 11:14, Tim Behrens wrote: > You need to merge your data from different diffusion directions > into a single 4D nifti file - you can use fslmerge to do this. > > T > On 15 Oct 2007, at 09:52, Patrick Hales wrote: > >> Hi >> >> This is the first time I have tried using FSL, and I am trying to >> get the DTIFit program to work. I have >> a series of diffusion weighted images, each of which are stored as >> individual fid files (created using >> Varian's VnmrJ software). There are seven images in total, one >> with b=0 and 6 with diffusion >> gradients applied in non co-linear directions. I have converted >> each of these into Nifti format, and >> stored them all in a directory. I used FSLview to create my binary >> mask, and have created text files >> with my gradient directions (all normalized to unit length), and >> my b-values (in units of s/mm2). In >> the DTIFit program, I have selected 'specify input files >> manually', and supplied it with the directory >> containing my nifti files, and the filenames of my gradients and b- >> values text files. >> >> When I press 'go', I get a number of error messages, such as: >> 'ERROR: nifti_image_read(....directory >> name....)can't open header file', 'ERROR: nifti_image_open >> (...):bad header info', 'Error: failed to open >> file', etc. I get the same errors when I try it with analyze files. >> >> Would you be able to tell me what I'm doing wrong? Thanks very much >> >> Patrick Hales ========================================================================= Date: Mon, 15 Oct 2007 11:45:04 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Dave Flitney <[log in to unmask]> Subject: Re: FSL on XGrid w/ XServe (now SGE) In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: multipart/alternative; boundary=Apple-Mail-1--610460582 --Apple-Mail-1--610460582 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Jeremy, On 13 Oct 2007, at 04:52, Jeremy Bronson wrote: > Hi Lokke, > > Thanks for the response. We have a very similar setup, and as > usual all users authenticate via LDAP/Kerberos to the XServe for > access to their accounts, shares and Kerberized services. This > works for Windows and Linux clients as well as Mac. The problem is > that SGE seems to explicitly state that the nodes in the grid must > have identical _local_ accounts. This makes a certain kind of > sense, it's the other way to solve the issue of permissions besides > using the 'nobody' user, but it doesn't seem like it scales well. You don't need local accounts. Our compute nodes only authenticate/ authorise against their LDAP server, i.e., they are setup to get account details from LDAP alone - no _local_ accounts bar the pre- installed system ones. Our users submit jobs under their own account id so SGE can give them a fair share of the system. We use a single submission host but, because all our desktop machines have a similar setup to the compute nodes, we could make them submission hosts too. Simply configure your nodes to use the same LDAP services and share file systems via NFS and all should be okay. > Imagine trying to administer a grid at a university or company > across buildings, networks, os types and versions, and ensuring SGE can deal with different OSs and architectures (we have PowerPC running MacOS and Sun AMD64 with Fedora Core in our cluster). > that every system has the same user accounts with the same IDs per > user - it's functionally impossible. The other method using > 'nobody' is used by XGrid, which allows for massively distributed > grid processing, a la SETI@Home and Stanford's XGrid projects. Any > user anywhere in the world can elect to become a grid node without > any special user accounts on their system. Yes SGE is unsuited to the SETI@HOME approach. SGE is not designed for ad hoc systems of that type. On the other hand writing programs which can exploit those types of mechanism is much harder: getting data/results to and from target machines; distribute the correct binaries; deal with some muppet who subscribes his/her shambles of a computer into your system and then reboots it every 5 mins. If you want to use such a system you will have to recode. Our scripts have been written to work with SGE but the job submission is abstracted out in the fsl_sub script do if you want to use some other system this is where you should customise things. Any mechanism should work so long as: binaries can be located in a predictable place; paths to the data are the same regardless of where the program runs. Of course if your chosen system needs to move data to/from the execution hosts then you have a much bigger problem on your hands. -- Cheers, Dave Dave Flitney, IT Manager Oxford Centre for Functional MRI of the Brain E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 URL: http://www.fmrib.ox.ac.uk/~flitney --Apple-Mail-1--610460582 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1 Jeremy,

On 13 = Oct 2007, at 04:52, Jeremy Bronson wrote:

Hi Lokke,


Thanks for the response.=A0= We have a very similar setup, and as usual all users = authenticate via LDAP/Kerberos to the XServe for access to their = accounts, shares and Kerberized services.=A0 This works for Windows and = Linux clients as well as Mac. The problem is that SGE seems to = explicitly state that the nodes in the grid must have identical _local_ = accounts.=A0 This makes a = certain kind of sense, it's the other way to solve the issue of = permissions besides using the 'nobody' user, but it doesn't seem like it = scales well.=A0

You don't need local accounts. = Our compute nodes only authenticate/authorise against their LDAP server, = i.e., they are setup to get account details from LDAP alone - no _local_ = accounts bar the pre-installed system ones. Our users submit jobs under = their own account id so SGE can give them a fair share of the system. We = use a single submission host but, because all our desktop machines have = a similar setup to the compute nodes, we could make them submission = hosts too.

Simply configure your nodes = to use the same LDAP services and share file systems via NFS and all = should be okay.

Imagine trying to administer = a grid at a university or company across buildings, networks, os types = and versions, and ensuring


SGE can = deal with different OSs and architectures (we have PowerPC running MacOS = and Sun AMD64 with Fedora Core in our cluster).

that = every system has the same user accounts with the same IDs per user - = it's functionally impossible.=A0 = The other method using 'nobody' is used by XGrid, which allows = for massively distributed grid processing, a la SETI@Home and Stanford's = XGrid projects.=A0 Any user = anywhere in the world can elect to become a grid node without any = special user accounts on their = system.


Yes SGE is unsuited to the = SETI@HOME approach. SGE is not designed for ad hoc systems of that = type.=A0

On = the other hand writing programs which can exploit those types of = mechanism is much harder: getting data/results to and from target = machines; distribute the correct binaries; deal with some muppet who = subscribes his/her shambles of a computer into your system and then = reboots it every 5 mins.

If you want to use such a = system you will have to recode. Our scripts have been written to work = with SGE but the job submission is abstracted out in the fsl_sub script = do if you want to use some other system this is where you should = customise things. Any mechanism should work so long as: binaries can be = located in a predictable place; paths to the data are the same = regardless of where the program runs. Of course if your chosen system = needs to move data to/from the execution hosts then you have a much = bigger problem on your hands.

=
Dave Flitney, = IT Manager
Oxford Centre for Functional MRI = of the Brain
E:[log in to unmask] = W:+44-1865-222713 F:+44-1865-222717

=

= --Apple-Mail-1--610460582-- ========================================================================= Date: Mon, 15 Oct 2007 13:00:17 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Tim Behrens <[log in to unmask]> Subject: Re: Problem with DTIFit In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit If you have avwmerge then you have an old version of FSL - get your sysadmin to upgrade. there may be many things that have gone wrong with data - The first thing to do is to make sure your files are named according to the bedpostx specifications, and then run bedpostx_datacheck on the directory - this will find any obvious problems. See here for instructions: http://www.fmrib.ox.ac.uk/fsl/fdt/fdt_bedpostx.html On 15 Oct 2007, at 11:36, Patrick Hales wrote: > Hi > > Thanks for that. The command 'fslmerge' isn't recognized for some > reason, however, I have been able to make my 4D data file using > 'avwmerge' instead. Unfortunately, when I try to process the data > in DTIFit, I now get the error message: "Error: child killed: bus > error". Would you be able to tell me how to fix this? > > Thanks very much for your help > > Patrick > > > > > On 15 Oct 2007, at 11:14, Tim Behrens wrote: > >> You need to merge your data from different diffusion directions >> into a single 4D nifti file - you can use fslmerge to do this. >> >> T >> On 15 Oct 2007, at 09:52, Patrick Hales wrote: >> >>> Hi >>> >>> This is the first time I have tried using FSL, and I am trying to >>> get the DTIFit program to work. I have >>> a series of diffusion weighted images, each of which are stored >>> as individual fid files (created using >>> Varian's VnmrJ software). There are seven images in total, one >>> with b=0 and 6 with diffusion >>> gradients applied in non co-linear directions. I have converted >>> each of these into Nifti format, and >>> stored them all in a directory. I used FSLview to create my >>> binary mask, and have created text files >>> with my gradient directions (all normalized to unit length), and >>> my b-values (in units of s/mm2). In >>> the DTIFit program, I have selected 'specify input files >>> manually', and supplied it with the directory >>> containing my nifti files, and the filenames of my gradients and >>> b-values text files. >>> >>> When I press 'go', I get a number of error messages, such as: >>> 'ERROR: nifti_image_read(....directory >>> name....)can't open header file', 'ERROR: nifti_image_open >>> (...):bad header info', 'Error: failed to open >>> file', etc. I get the same errors when I try it with analyze files. >>> >>> Would you be able to tell me what I'm doing wrong? Thanks very much >>> >>> Patrick Hales ========================================================================= Date: Mon, 15 Oct 2007 08:00:50 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matt Glasser <[log in to unmask]> Organization: ma-tea.com Subject: Re: Problem with DTIFit In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Hi, Fslmerge is in the latest version of FSL whereas avwmerge is in earlier versions. You could navigate to the data directory with the commandline and try running bedpostx_datacheck and post that output here. Peace, Matt. -----Original Message----- From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Patrick Hales Sent: Monday, October 15, 2007 6:36 AM To: [log in to unmask] Subject: Re: [FSL] Problem with DTIFit Hi Thanks for that. The command 'fslmerge' isn't recognized for some reason, however, I have been able to make my 4D data file using 'avwmerge' instead. Unfortunately, when I try to process the data in DTIFit, I now get the error message: "Error: child killed: bus error". Would you be able to tell me how to fix this? Thanks very much for your help Patrick On 15 Oct 2007, at 11:14, Tim Behrens wrote: > You need to merge your data from different diffusion directions > into a single 4D nifti file - you can use fslmerge to do this. > > T > On 15 Oct 2007, at 09:52, Patrick Hales wrote: > >> Hi >> >> This is the first time I have tried using FSL, and I am trying to >> get the DTIFit program to work. I have >> a series of diffusion weighted images, each of which are stored as >> individual fid files (created using >> Varian's VnmrJ software). There are seven images in total, one >> with b=0 and 6 with diffusion >> gradients applied in non co-linear directions. I have converted >> each of these into Nifti format, and >> stored them all in a directory. I used FSLview to create my binary >> mask, and have created text files >> with my gradient directions (all normalized to unit length), and >> my b-values (in units of s/mm2). In >> the DTIFit program, I have selected 'specify input files >> manually', and supplied it with the directory >> containing my nifti files, and the filenames of my gradients and b- >> values text files. >> >> When I press 'go', I get a number of error messages, such as: >> 'ERROR: nifti_image_read(....directory >> name....)can't open header file', 'ERROR: nifti_image_open >> (...):bad header info', 'Error: failed to open >> file', etc. I get the same errors when I try it with analyze files. >> >> Would you be able to tell me what I'm doing wrong? Thanks very much >> >> Patrick Hales ========================================================================= Date: Mon, 15 Oct 2007 08:00:02 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Marc Dubin <[log in to unmask]> Subject: Re: mean_FA_skeleton and brain misalignment In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_21544_31278454.1192449602568" ------=_Part_21544_31278454.1192449602568 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Thanks.....Let me try your fslview overlay suggestion and I will let you know. If I am still having trouble, I will upload the tarfile that you requested. Thanks again! On 10/15/07, Steve Smith <[log in to unmask]> wrote: > > Hi - I'm suprised that you get this with the -T option. If you load > in all_FA first into FSLView, and then overlay the FMRIB58_FA > (threshold this, change its colour to red-yellow, and make it a > little transparent) and then move through the timepoints, do all > subjects behave in a similar way to what you've described? > > We can have a look and advise on what's happening; > Please upload the files in a single compressed tarfile to > http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi > and then email me the upload ID. > > Cheers. > > > > On 15 Oct 2007, at 02:36, Marc Dubin wrote: > > > Hello FSL Users, > > > > Using the TBSS tools I have been carrying out the tbss_2_reg step > > with the -T option (FMRIB58_FA). I > > then go through all of the remaining steps of TBSS and then when I > > visualise with: > > > > fslview MNI152 mean_FA_skeleton -l Green -b 2000,8000 tbss_tstat1 - > > l Red-Yellow -b 3,6 > > tbss_tstat2 -l Blue-Lightblue -b 3,6 > > > > the mean_FA_skeleton significantly overshoots the brain, > > terminating in or even beyond the skull > > (the overall extent of the skeleton is significantly larger that > > the brain). > > > > Any ideas about what might be going on here would be greatly > > appreciated! > > > > Thanks, > > Marc > > > ------------------------------------------------------------------------ > --- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > ------------------------------------------------------------------------ > --- > -- Marc Dubin, MD PhD Resident in Psychiatry - PGY4 Weill Cornell Medical College 212-746-3764 (w) 646-831-8886 (c) ------=_Part_21544_31278454.1192449602568 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Thanks.....Let me try your fslview overlay suggestion and I will let you know. If I am still having trouble, I will upload the tarfile that you requested.

Thanks again!

On 10/15/07, Steve Smith <[log in to unmask]> wrote:
Hi - I'm suprised that you get this with the -T option. If you load
in all_FA first into FSLView, and then overlay the FMRIB58_FA
(threshold this, change its colour to red-yellow, and make it a
little transparent) and then move through the timepoints, do all
subjects behave in a similar way to what you've described?

We can have a look and advise on what's happening;
Please upload the files in a single compressed tarfile to
http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi
and then email me the upload ID.

Cheers.



On 15 Oct 2007, at 02:36, Marc Dubin wrote:

> Hello FSL Users,
>
> Using the TBSS tools I have been carrying out the tbss_2_reg step
> with the -T option (FMRIB58_FA). I
> then go through all of the remaining steps of TBSS and then when I
> visualise with:
>
> fslview MNI152 mean_FA_skeleton -l Green -b 2000,8000 tbss_tstat1 -
> l Red-Yellow -b 3,6
> tbss_tstat2 -l Blue-Lightblue -b 3,6
>
> the mean_FA_skeleton significantly overshoots the brain,
> terminating in or even beyond the skull
> (the overall extent of the skeleton is significantly larger that
> the brain).
>
> Any ideas about what might be going on here would be greatly
> appreciated!
>
> Thanks,
> Marc


------------------------------------------------------------ ------------
---
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director,  Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222726  (fax 222717)
[log in to unmask]    http://www.fmrib.ox.ac.uk/~steve
------------------------------------------------------------------------
---



--
Marc Dubin, MD PhD
Resident in Psychiatry - PGY4
Weill Cornell Medical College

212-746-3764 (w) 646-831-8886 (c)
------=_Part_21544_31278454.1192449602568-- ========================================================================= Date: Mon, 15 Oct 2007 08:03:42 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matt Glasser <[log in to unmask]> Organization: ma-tea.com Subject: Re: DTI processing In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Try using fslmaths -mul 1000 to multiply the data by 1000. This should display better in the 3D viewer. I haven't been able to get it to work with low numbers either. Peace, Matt. -----Original Message----- From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Markus Gschwind Sent: Monday, October 15, 2007 6:05 AM To: [log in to unmask] Subject: Re: [FSL] DTI processing Hi Steve! Thanks for the answers. Quoting Steve Smith <[log in to unmask]>: >> 1. >> I can open FA data in FSLveiw 2D nicely, but I cannot open any >> FA-data (produced in FSL3 on Cygwin) in FSLview 3D. That means: It >> atually works but I just don't see anything! Is that normal? > > Have you tried setting the FSLView display range to 0:1? Yes exactly. But I dont see anything. Acutally in the terminal it tells me: "ERROR: In /home/duncan/VTK/Graphics/vtkPolyDataNormals.cxx, line 94 vtkPolyDataNormals (0xa2fccb8): No data to generate normals for!" Is that a compatibility problem of FSL 3 and FSL 4? Thanks again, cheers, Markus >> -- >> >> Markus Gschwind, M.D. >> Laboratory for Neurology and Imaging of Cognition >> Dept of Neurosciences >> University Medical Center (CMU) >> 1 Michel-Servet - 1211 GENEVA - CH >> >> Tel 0041 (0) 22 379 5324 >> Fax 0041 (0) 22 379 5402 >> email: [log in to unmask] >> http://labnic.unige.ch > > > --------------------------------------------------------------------------- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > --------------------------------------------------------------------------- -- Markus Gschwind, M.D. Laboratory for Neurology and Imaging of Cognition Dept of Neurosciences University Medical Center (CMU) 1 Michel-Servet - 1211 GENEVA - CH Tel 0041 (0) 22 379 5324 Fax 0041 (0) 22 379 5402 email: [log in to unmask] http://labnic.unige.ch ========================================================================= Date: Mon, 15 Oct 2007 14:28:25 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Tim Behrens <[log in to unmask]> Subject: Re: Problem with DTIFit In-Reply-To: [log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Although - if you have the old version of FSL, it will be called bedpost_datacheck! T On 15 Oct 2007, at 13:00, Matt Glasser wrote: > Hi, > > Fslmerge is in the latest version of FSL whereas avwmerge is in > earlier > versions. You could navigate to the data directory with the > commandline and > try running bedpostx_datacheck and post that output here. > > Peace, > > Matt. > > -----Original Message----- > From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On > Behalf > Of Patrick Hales > Sent: Monday, October 15, 2007 6:36 AM > To: [log in to unmask] > Subject: Re: [FSL] Problem with DTIFit > > Hi > > Thanks for that. The command 'fslmerge' isn't recognized for some > reason, however, I have been able to make my 4D data file using > 'avwmerge' instead. Unfortunately, when I try to process the data in > DTIFit, I now get the error message: "Error: child killed: bus > error". Would you be able to tell me how to fix this? > > Thanks very much for your help > > Patrick > > > > > On 15 Oct 2007, at 11:14, Tim Behrens wrote: > >> You need to merge your data from different diffusion directions >> into a single 4D nifti file - you can use fslmerge to do this. >> >> T >> On 15 Oct 2007, at 09:52, Patrick Hales wrote: >> >>> Hi >>> >>> This is the first time I have tried using FSL, and I am trying to >>> get the DTIFit program to work. I have >>> a series of diffusion weighted images, each of which are stored as >>> individual fid files (created using >>> Varian's VnmrJ software). There are seven images in total, one >>> with b=0 and 6 with diffusion >>> gradients applied in non co-linear directions. I have converted >>> each of these into Nifti format, and >>> stored them all in a directory. I used FSLview to create my binary >>> mask, and have created text files >>> with my gradient directions (all normalized to unit length), and >>> my b-values (in units of s/mm2). In >>> the DTIFit program, I have selected 'specify input files >>> manually', and supplied it with the directory >>> containing my nifti files, and the filenames of my gradients and b- >>> values text files. >>> >>> When I press 'go', I get a number of error messages, such as: >>> 'ERROR: nifti_image_read(....directory >>> name....)can't open header file', 'ERROR: nifti_image_open >>> (...):bad header info', 'Error: failed to open >>> file', etc. I get the same errors when I try it with analyze files. >>> >>> Would you be able to tell me what I'm doing wrong? Thanks very much >>> >>> Patrick Hales ========================================================================= Date: Mon, 15 Oct 2007 09:57:08 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matt Glasser <[log in to unmask]> Organization: ma-tea.com Subject: Re: DTI processing MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Hi, There isn't currently a good way to display probtrackx results in the 3D mode. You can try clicking the "I" button and setting the image type to Mask/Label or Statistic, as either of these will be displayed, and you can see which you like better. I always have to reload the 3D window if I want to show these things. You should read the page about the 3D viewer to learn about all of the features currently available: http://www.fmrib.ox.ac.uk/fsl/fslview/3D.html Peace, Matt. -----Original Message----- From: [log in to unmask] [mailto:[log in to unmask]] Sent: Monday, October 15, 2007 9:20 AM To: Matt Glasser Subject: Re: [FSL] DTI processing Hi Matt! Thanks yeahh! It is working. But already the next problem: I saved some fibers in a nii.gz file. Those aswell I can visualise nicely in fslview 2D (red) BUT in 3D I dont see anything -neither with the opriginal nor with the (-mul 1000) one (cf attached screen shot). Maybe you know that problem also? ;-) Cheers, Markus Quoting Matt Glasser <[log in to unmask]>: > Try using fslmaths -mul 1000 to multiply the > data by 1000. This should display better in the 3D viewer. I haven't been > able to get it to work with low numbers either. > > Peace, > > Matt. > > -----Original Message----- > From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf > Of Markus Gschwind > Sent: Monday, October 15, 2007 6:05 AM > To: [log in to unmask] > Subject: Re: [FSL] DTI processing > > Hi Steve! > Thanks for the answers. > > Quoting Steve Smith <[log in to unmask]>: >>> 1. >>> I can open FA data in FSLveiw 2D nicely, but I cannot open any >>> FA-data (produced in FSL3 on Cygwin) in FSLview 3D. That means: It >>> atually works but I just don't see anything! Is that normal? >> >> Have you tried setting the FSLView display range to 0:1? > > Yes exactly. But I dont see anything. Acutally in the terminal it tells me: > > "ERROR: In /home/duncan/VTK/Graphics/vtkPolyDataNormals.cxx, line 94 > vtkPolyDataNormals (0xa2fccb8): No data to generate normals for!" > > Is that a compatibility problem of FSL 3 and FSL 4? > > Thanks again, cheers, > Markus > > >>> -- >>> >>> Markus Gschwind, M.D. >>> Laboratory for Neurology and Imaging of Cognition >>> Dept of Neurosciences >>> University Medical Center (CMU) >>> 1 Michel-Servet - 1211 GENEVA - CH >>> >>> Tel 0041 (0) 22 379 5324 >>> Fax 0041 (0) 22 379 5402 >>> email: [log in to unmask] >>> http://labnic.unige.ch >> >> >> > --------------------------------------------------------------------------- >> Stephen M. Smith, Professor of Biomedical Engineering >> Associate Director, Oxford University FMRIB Centre >> >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >> +44 (0) 1865 222726 (fax 222717) >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >> > --------------------------------------------------------------------------- > > > > -- > > Markus Gschwind, M.D. > Laboratory for Neurology and Imaging of Cognition > Dept of Neurosciences > University Medical Center (CMU) > 1 Michel-Servet - 1211 GENEVA - CH > > Tel 0041 (0) 22 379 5324 > Fax 0041 (0) 22 379 5402 > email: [log in to unmask] > http://labnic.unige.ch > -- Markus Gschwind, M.D. Laboratory for Neurology and Imaging of Cognition Dept of Neurosciences University Medical Center (CMU) 1 Michel-Servet - 1211 GENEVA - CH Tel 0041 (0) 22 379 5324 Fax 0041 (0) 22 379 5402 email: [log in to unmask] http://labnic.unige.ch ========================================================================= Date: Mon, 15 Oct 2007 15:53:29 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Patrick Hales <[log in to unmask]> Subject: Re: Problem with DTIFit In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi Thanks again for the help. I've tried running bedpost_datacheck on my data (after renaming everything to the appropriate filenames). The output I get is: bedpost_data/data data_type INT16 dim1 128 dim2 128 dim3 7 dim4 1 datatype 4 pixdim1 0.2343750000 pixdim2 0.2343750000 pixdim3 1.0000000000 pixdim4 1.0000000000 cal_max 25500.0000 cal_min 40.0000 file_type NIFTI-1+ bedpost_data/nodif does not exist bedpost_data/nodif_brain_mask data_type INT16 dim1 128 dim2 128 dim3 7 dim4 1 datatype 4 pixdim1 0.2343750000 pixdim2 0.2343750000 pixdim3 1.0000000000 pixdim4 1.0000000000 cal_max 4.0000 cal_min 0.0000 file_type NIFTI-1+ num lines in bedpost_data/bvals 1 num words in bedpost_data/bvals 7 num lines in bedpost_data/bvecs 3 num words in bedpost_data/bvecs 21 number of elements in bvals is not equal to number of vols in data number of elements per line in bvecs is not equal to number of vols in data I don't understand what's wrong with my bvals and bvecs files. My 'data' file should now be made up of 7 volumes, the 1st with b=0 and the other 6 with diffusion weighting in various directions. My bvals file looks like this: 0 1025 1048 826 1003 582 826 and my bvecs file looks like this: 0 1 0 1 0 1 -1 0 1 1 0 1 -1 0 0 0 1 1 -1 0 1 i.e. the gradients directions have been normalized to unit length, and the vectors run in columns. Could you tell me what I'm doing wrong with these files? - I've read the help and it sounds like I have set them up correctly, but I'm obviously doing something wrong. Thanks very much Patrick On 15 Oct 2007, at 14:28, Tim Behrens wrote: > Although - if you have the old version of FSL, it will be called > bedpost_datacheck! > T > On 15 Oct 2007, at 13:00, Matt Glasser wrote: > >> Hi, >> >> Fslmerge is in the latest version of FSL whereas avwmerge is in >> earlier >> versions. You could navigate to the data directory with the >> commandline and >> try running bedpostx_datacheck and post that output here. >> >> Peace, >> >> Matt. >> >> -----Original Message----- >> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] >> On Behalf >> Of Patrick Hales >> Sent: Monday, October 15, 2007 6:36 AM >> To: [log in to unmask] >> Subject: Re: [FSL] Problem with DTIFit >> >> Hi >> >> Thanks for that. The command 'fslmerge' isn't recognized for some >> reason, however, I have been able to make my 4D data file using >> 'avwmerge' instead. Unfortunately, when I try to process the data in >> DTIFit, I now get the error message: "Error: child killed: bus >> error". Would you be able to tell me how to fix this? >> >> Thanks very much for your help >> >> Patrick >> >> >> >> >> On 15 Oct 2007, at 11:14, Tim Behrens wrote: >> >>> You need to merge your data from different diffusion directions >>> into a single 4D nifti file - you can use fslmerge to do this. >>> >>> T >>> On 15 Oct 2007, at 09:52, Patrick Hales wrote: >>> >>>> Hi >>>> >>>> This is the first time I have tried using FSL, and I am trying to >>>> get the DTIFit program to work. I have >>>> a series of diffusion weighted images, each of which are stored as >>>> individual fid files (created using >>>> Varian's VnmrJ software). There are seven images in total, one >>>> with b=0 and 6 with diffusion >>>> gradients applied in non co-linear directions. I have converted >>>> each of these into Nifti format, and >>>> stored them all in a directory. I used FSLview to create my binary >>>> mask, and have created text files >>>> with my gradient directions (all normalized to unit length), and >>>> my b-values (in units of s/mm2). In >>>> the DTIFit program, I have selected 'specify input files >>>> manually', and supplied it with the directory >>>> containing my nifti files, and the filenames of my gradients and b- >>>> values text files. >>>> >>>> When I press 'go', I get a number of error messages, such as: >>>> 'ERROR: nifti_image_read(....directory >>>> name....)can't open header file', 'ERROR: nifti_image_open >>>> (...):bad header info', 'Error: failed to open >>>> file', etc. I get the same errors when I try it with analyze files. >>>> >>>> Would you be able to tell me what I'm doing wrong? Thanks very much >>>> >>>> Patrick Hales ========================================================================= Date: Mon, 15 Oct 2007 15:57:07 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Tim Behrens <[log in to unmask]> Subject: Re: Problem with DTIFit In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: multipart/alternative; boundary=Apple-Mail-4--595336703 --Apple-Mail-4--595336703 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; format=flowed Your data file has 7 **slices** and 1 volume. you need to use fslmerge with the -t option T On 15 Oct 2007, at 15:53, Patrick Hales wrote: > > bedpost_data/data > data_type INT16 > dim1 128 > dim2 128 > dim3 7 > dim4 1 --Apple-Mail-4--595336703 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1 Your data file has 7 **slices** = and 1 volume.

you = need to use fslmerge with the -t option

T
On 15 Oct = 2007, at 15:53, Patrick Hales wrote:


bedpost_data/data

data_type=A0 =A0 =A0 = INT16

dim1 =A0 =A0 =A0 =A0 =A0 = 128

dim2 = =A0 =A0 =A0 =A0 =A0 = 128

dim3 = =A0 =A0 =A0 =A0 =A0 = 7

dim4 = =A0 =A0 =A0 =A0 =A0 = 1


= --Apple-Mail-4--595336703-- ========================================================================= Date: Mon, 15 Oct 2007 16:13:59 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Tim Behrens <[log in to unmask]> Subject: Re: Problem with DTIFit In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit PS - > 0 1 0 1 0 1 -1 > 0 1 1 0 1 -1 0 > 0 0 1 1 -1 0 1 > > i.e. the gradients directions have been normalized to unit length, These are not unit length. They have a length of sqrt(2) This does not matter for the new version of FDT, but probably does matter for the old version you are working with T On 15 Oct 2007, at 15:53, Patrick Hales wrote: > Hi > > Thanks again for the help. I've tried running bedpost_datacheck on > my data (after renaming everything to the appropriate filenames). > The output I get is: > > bedpost_data/data > data_type INT16 > dim1 128 > dim2 128 > dim3 7 > dim4 1 > datatype 4 > pixdim1 0.2343750000 > pixdim2 0.2343750000 > pixdim3 1.0000000000 > pixdim4 1.0000000000 > cal_max 25500.0000 > cal_min 40.0000 > file_type NIFTI-1+ > > bedpost_data/nodif does not exist > bedpost_data/nodif_brain_mask > data_type INT16 > dim1 128 > dim2 128 > dim3 7 > dim4 1 > datatype 4 > pixdim1 0.2343750000 > pixdim2 0.2343750000 > pixdim3 1.0000000000 > pixdim4 1.0000000000 > cal_max 4.0000 > cal_min 0.0000 > file_type NIFTI-1+ > > num lines in bedpost_data/bvals > 1 > num words in bedpost_data/bvals > 7 > num lines in bedpost_data/bvecs > 3 > num words in bedpost_data/bvecs > 21 > number of elements in bvals is not equal to number of vols in data > number of elements per line in bvecs is not equal to number of vols > in data > > I don't understand what's wrong with my bvals and bvecs files. My > 'data' file should now be made up of 7 volumes, the 1st with b=0 > and the other 6 with diffusion weighting in various directions. My > bvals file looks like this: > > 0 1025 1048 826 1003 582 826 > > and my bvecs file looks like this: > > 0 1 0 1 0 1 -1 > 0 1 1 0 1 -1 0 > 0 0 1 1 -1 0 1 > > i.e. the gradients directions have been normalized to unit length, > and the vectors run in columns. Could you tell me what I'm doing > wrong with these files? - I've read the help and it sounds like I > have set them up correctly, but I'm obviously doing something wrong. > > Thanks very much > > Patrick > > > > > > > On 15 Oct 2007, at 14:28, Tim Behrens wrote: > >> Although - if you have the old version of FSL, it will be called >> bedpost_datacheck! >> T >> On 15 Oct 2007, at 13:00, Matt Glasser wrote: >> >>> Hi, >>> >>> Fslmerge is in the latest version of FSL whereas avwmerge is in >>> earlier >>> versions. You could navigate to the data directory with the >>> commandline and >>> try running bedpostx_datacheck and post that output here. >>> >>> Peace, >>> >>> Matt. >>> >>> -----Original Message----- >>> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] >>> On Behalf >>> Of Patrick Hales >>> Sent: Monday, October 15, 2007 6:36 AM >>> To: [log in to unmask] >>> Subject: Re: [FSL] Problem with DTIFit >>> >>> Hi >>> >>> Thanks for that. The command 'fslmerge' isn't recognized for some >>> reason, however, I have been able to make my 4D data file using >>> 'avwmerge' instead. Unfortunately, when I try to process the data in >>> DTIFit, I now get the error message: "Error: child killed: bus >>> error". Would you be able to tell me how to fix this? >>> >>> Thanks very much for your help >>> >>> Patrick >>> >>> >>> >>> >>> On 15 Oct 2007, at 11:14, Tim Behrens wrote: >>> >>>> You need to merge your data from different diffusion directions >>>> into a single 4D nifti file - you can use fslmerge to do this. >>>> >>>> T >>>> On 15 Oct 2007, at 09:52, Patrick Hales wrote: >>>> >>>>> Hi >>>>> >>>>> This is the first time I have tried using FSL, and I am trying to >>>>> get the DTIFit program to work. I have >>>>> a series of diffusion weighted images, each of which are stored as >>>>> individual fid files (created using >>>>> Varian's VnmrJ software). There are seven images in total, one >>>>> with b=0 and 6 with diffusion >>>>> gradients applied in non co-linear directions. I have converted >>>>> each of these into Nifti format, and >>>>> stored them all in a directory. I used FSLview to create my binary >>>>> mask, and have created text files >>>>> with my gradient directions (all normalized to unit length), and >>>>> my b-values (in units of s/mm2). In >>>>> the DTIFit program, I have selected 'specify input files >>>>> manually', and supplied it with the directory >>>>> containing my nifti files, and the filenames of my gradients >>>>> and b- >>>>> values text files. >>>>> >>>>> When I press 'go', I get a number of error messages, such as: >>>>> 'ERROR: nifti_image_read(....directory >>>>> name....)can't open header file', 'ERROR: nifti_image_open >>>>> (...):bad header info', 'Error: failed to open >>>>> file', etc. I get the same errors when I try it with analyze >>>>> files. >>>>> >>>>> Would you be able to tell me what I'm doing wrong? Thanks very >>>>> much >>>>> >>>>> Patrick Hales ========================================================================= Date: Mon, 15 Oct 2007 17:12:10 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Markus Gschwind <[log in to unmask]> Subject: Re: DTI processing In-Reply-To: [log in to unmask]> MIME-version: 1.0 Content-transfer-encoding: 7BIT Content-disposition: inline Content-type: text/plain; charset=ISO-8859-1; DelSp=Yes; format=flowed Hi! Thanks again. Yes infact, I have already read that page and I tried the i button and I still couldnt see anything ;-) and then I wrote that mail. When I choose another dysplay mode (statistic or mask or DTI...) I cannot go back anymore and the fslviewer gets stuck. But if you tell me that with "mask" it should work... well, I remember also in the FSL course that we did it and it worked. It was a kind of dots-trace throw the brain. Looked nice ;-) Thanks again, peace also, Markus Quoting Matt Glasser <[log in to unmask]>: > Hi, > > There isn't currently a good way to display probtrackx results in the 3D > mode. You can try clicking the "I" button and setting the image type to > Mask/Label or Statistic, as either of these will be displayed, and you can > see which you like better. I always have to reload the 3D window if I want > to show these things. You should read the page about the 3D viewer to learn > about all of the features currently available: > http://www.fmrib.ox.ac.uk/fsl/fslview/3D.html > > Peace, > > Matt. > > -----Original Message----- > From: [log in to unmask] > [mailto:[log in to unmask]] > Sent: Monday, October 15, 2007 9:20 AM > To: Matt Glasser > Subject: Re: [FSL] DTI processing > > Hi Matt! > Thanks yeahh! It is working. > But already the next problem: I saved some fibers in a nii.gz file. > Those aswell I can visualise nicely in fslview 2D (red) BUT in 3D I > dont see anything -neither with the opriginal nor with the (-mul 1000) > one (cf attached screen shot). > Maybe you know that problem also? ;-) > > Cheers, > Markus > > > Quoting Matt Glasser <[log in to unmask]>: > >> Try using fslmaths -mul 1000 to multiply the >> data by 1000. This should display better in the 3D viewer. I haven't > been >> able to get it to work with low numbers either. >> >> Peace, >> >> Matt. >> >> -----Original Message----- >> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf >> Of Markus Gschwind >> Sent: Monday, October 15, 2007 6:05 AM >> To: [log in to unmask] >> Subject: Re: [FSL] DTI processing >> >> Hi Steve! >> Thanks for the answers. >> >> Quoting Steve Smith <[log in to unmask]>: >>>> 1. >>>> I can open FA data in FSLveiw 2D nicely, but I cannot open any >>>> FA-data (produced in FSL3 on Cygwin) in FSLview 3D. That means: It >>>> atually works but I just don't see anything! Is that normal? >>> >>> Have you tried setting the FSLView display range to 0:1? >> >> Yes exactly. But I dont see anything. Acutally in the terminal it tells > me: >> >> "ERROR: In /home/duncan/VTK/Graphics/vtkPolyDataNormals.cxx, line 94 >> vtkPolyDataNormals (0xa2fccb8): No data to generate normals for!" >> >> Is that a compatibility problem of FSL 3 and FSL 4? >> >> Thanks again, cheers, >> Markus >> >> >>>> -- >>>> >>>> Markus Gschwind, M.D. >>>> Laboratory for Neurology and Imaging of Cognition >>>> Dept of Neurosciences >>>> University Medical Center (CMU) >>>> 1 Michel-Servet - 1211 GENEVA - CH >>>> >>>> Tel 0041 (0) 22 379 5324 >>>> Fax 0041 (0) 22 379 5402 >>>> email: [log in to unmask] >>>> http://labnic.unige.ch >>> >>> >>> >> > --------------------------------------------------------------------------- >>> Stephen M. Smith, Professor of Biomedical Engineering >>> Associate Director, Oxford University FMRIB Centre >>> >>> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >>> +44 (0) 1865 222726 (fax 222717) >>> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >>> >> > --------------------------------------------------------------------------- >> >> >> >> -- >> >> Markus Gschwind, M.D. >> Laboratory for Neurology and Imaging of Cognition >> Dept of Neurosciences >> University Medical Center (CMU) >> 1 Michel-Servet - 1211 GENEVA - CH >> >> Tel 0041 (0) 22 379 5324 >> Fax 0041 (0) 22 379 5402 >> email: [log in to unmask] >> http://labnic.unige.ch >> > > > > -- > > Markus Gschwind, M.D. > Laboratory for Neurology and Imaging of Cognition > Dept of Neurosciences > University Medical Center (CMU) > 1 Michel-Servet - 1211 GENEVA - CH > > Tel 0041 (0) 22 379 5324 > Fax 0041 (0) 22 379 5402 > email: [log in to unmask] > http://labnic.unige.ch > -- Markus Gschwind, M.D. Laboratory for Neurology and Imaging of Cognition Dept of Neurosciences University Medical Center (CMU) 1 Michel-Servet - 1211 GENEVA - CH Tel 0041 (0) 22 379 5324 Fax 0041 (0) 22 379 5402 email: [log in to unmask] http://labnic.unige.ch ========================================================================= Date: Mon, 15 Oct 2007 11:22:08 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Najmeh Khalili M." <[log in to unmask]> Subject: DTI: noise estimation MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi, Is there a way to get the residuals of eddy_correction or is there any way to estimate a metric of suseptibility distortions present in DTI? More generally, I need to model the imaging noise in the between-subject DTI analysis and I wonder if you have any tools < or philosophy > that address this problem. Thank you, Naj ========================================================================= Date: Mon, 15 Oct 2007 13:25:53 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: =?iso-8859-1?q?Gwena=EBlle=20DOUAUD?= <[log in to unmask]> Subject: RE : [FSL] mean_FA_skeleton and brain misalignment In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Hi Marc, could you please specify what you mean by "all of the remaining steps of TBSS", presumably tbss_3_postreg -S ans then tbss_4_prestats 2000? Could you also please provide us a snapshot of your mean_FA_skeleton overlaid onto the MNI152? Thanks, Gwenaelle --- Marc Dubin <[log in to unmask]> a =E9crit : > Hello FSL Users, >=20 > Using the TBSS tools I have been carrying out the > tbss_2_reg step with the -T option (FMRIB58_FA). I=20 > then go through all of the remaining steps of TBSS > and then when I visualise with: >=20 > fslview MNI152 mean_FA_skeleton -l Green -b > 2000,8000 tbss_tstat1 -l Red-Yellow -b 3,6=20 > tbss_tstat2 -l Blue-Lightblue -b 3,6 >=20 > the mean_FA_skeleton significantly overshoots the > brain, terminating in or even beyond the skull=20 > (the overall extent of the skeleton is significantly > larger that the brain). >=20 > Any ideas about what might be going on here would be > greatly appreciated! >=20 > Thanks, > Marc >=20 ___________________________________________________________________= __________=20 Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! M= ail=20 ========================================================================= Date: Mon, 15 Oct 2007 16:06:38 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Swati Rane <[log in to unmask]> Subject: Bedpost error (FSL 3.3) Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Hello, I am still using FSL3.3. When I run bedpost, it processes for a while and= then gives the following error: /usr/local/fsl/bin/bedpost: line 293: syntax error; unexpected end of fil= e Could you clarify why I get that error? Thanks! Swati ========================================================================= Date: Mon, 15 Oct 2007 16:54:12 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Tim Behrens <[log in to unmask]> Subject: Re: Bedpost error (FSL 3.3) In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Everyone used to get it - it is nothing to worry about. In fact it probably means everything has run ok! It is a feature that has been removed from the latest version! T On Mon, 15 Oct 2007, Swati Rane wrote: > Hello, > > I am still using FSL3.3. When I run bedpost, it processes for a while and > then gives the following error: > > /usr/local/fsl/bin/bedpost: line 293: syntax error; unexpected end of file > > Could you clarify why I get that error? > > Thanks! > > Swati > -- ------------------------------------------------------------------------------- Tim Behrens Centre for Functional MRI of the Brain The John Radcliffe Hospital Headley Way Oxford OX3 9DU Oxford University Work 01865 222782 Mobile 07980 884537 ------------------------------------------------------------------------------- ========================================================================= Date: Mon, 15 Oct 2007 16:59:10 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Dave Flitney <[log in to unmask]> Subject: Re: DTI processing In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: multipart/alternative; boundary=Apple-Mail-1--591614522 --Apple-Mail-1--591614522 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed On 15 Oct 2007, at 11:04, Markus Gschwind wrote: > Hi Steve! > Thanks for the answers. > > Quoting Steve Smith <[log in to unmask]>: >>> 1. >>> I can open FA data in FSLveiw 2D nicely, but I cannot open any >>> FA-data (produced in FSL3 on Cygwin) in FSLview 3D. That means: >>> It atually works but I just don't see anything! Is that normal? >> >> Have you tried setting the FSLView display range to 0:1? > > Yes exactly. But I dont see anything. Acutally in the terminal it > tells me: > > "ERROR: In /home/duncan/VTK/Graphics/vtkPolyDataNormals.cxx, line 94 > vtkPolyDataNormals (0xa2fccb8): No data to generate normals for!" Are you talking about 3D view or all views? Can you see your FA data in 2D view? There is no specific mode for showing FA data in 3D. 3D view will do a surface for the main image - threshold it with the box at the top of the display. It will also: blend in stats images and show an iso- surface for them; show the surface of mask images. If you want to view DTI data then perhaps you could try thresholding it appropriately (i.e., to isolate regions of similar FA) then setting it to Mask/Label in the image info dialog then switch to 3D. > Is that a compatibility problem of FSL 3 and FSL 4? > > Thanks again, cheers, > Markus > > >>> -- >>> >>> Markus Gschwind, M.D. >>> Laboratory for Neurology and Imaging of Cognition >>> Dept of Neurosciences >>> University Medical Center (CMU) >>> 1 Michel-Servet - 1211 GENEVA - CH >>> >>> Tel 0041 (0) 22 379 5324 >>> Fax 0041 (0) 22 379 5402 >>> email: [log in to unmask] >>> http://labnic.unige.ch >> >> >> --------------------------------------------------------------------- >> ------ >> Stephen M. Smith, Professor of Biomedical Engineering >> Associate Director, Oxford University FMRIB Centre >> >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >> +44 (0) 1865 222726 (fax 222717) >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >> --------------------------------------------------------------------- >> ------ > > > > -- > > Markus Gschwind, M.D. > Laboratory for Neurology and Imaging of Cognition > Dept of Neurosciences > University Medical Center (CMU) > 1 Michel-Servet - 1211 GENEVA - CH > > Tel 0041 (0) 22 379 5324 > Fax 0041 (0) 22 379 5402 > email: [log in to unmask] > http://labnic.unige.ch -- Cheers, Dave Dave Flitney, IT Manager Oxford Centre for Functional MRI of the Brain E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 URL: http://www.fmrib.ox.ac.uk/~flitney --Apple-Mail-1--591614522 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=ISO-8859-1

On 15 = Oct 2007, at 11:04, Markus Gschwind wrote:

Hi Steve!
Thanks for = the answers.

Quoting Steve Smith <[log in to unmask]>:
=
1.
I can open FA data in = FSLveiw 2D nicely, but I cannot open any=A0 FA-data (produced in FSL3 on = Cygwin) in FSLview 3D. That means: It=A0 atually works but I just = don't see anything! Is that normal?

Have you = tried setting the FSLView display range to 0:1?

Yes = exactly. But I dont see anything. Acutally in the terminal it tells = me:

"ERROR: In = /home/duncan/VTK/Graphics/vtkPolyDataNormals.cxx, line 94
vtkPolyDataNormals (0xa2fccb8): No data to generate = normals for!"

Are you talking about 3D view or = all views? Can you see your FA data in 2D view?=A0

There is no specific mode = for showing FA data in 3D. 3D view will do a surface for the main image = - threshold it with the box at the top of the display. It will also: = blend in stats images and show an iso-surface for them; show the surface = of mask images. If you want to view DTI data then perhaps you could try = thresholding it appropriately (i.e., to isolate regions of similar FA) = then setting it to Mask/Label in the image info dialog then switch to = 3D.

Is that a = compatibility problem of FSL 3 and FSL 4?

Thanks again, = cheers,
Markus


--=A0

Markus = Gschwind, M.D.
Laboratory for Neurology and = Imaging of Cognition
Dept of = Neurosciences
University Medical Center = (CMU)
1 Michel-Servet - 1211 GENEVA - = CH

Tel 0041 (0) 22 379 5324
Fax 0041 (0) 22 379 5402
=


Stephen M. Smith, Professor of = Biomedical Engineering
Associate Director,=A0 Oxford University FMRIB = Centre

FMRIB, JR Hospital, Headington, Oxford=A0 OX3 9DU, UK
+44 (0) 1865 222726=A0 (fax 222717)
[log in to unmask]=A0 =A0 = http://www.fmrib.ox.ac.uk/~steve



--=A0

Markus = Gschwind, M.D.
Laboratory for Neurology and = Imaging of Cognition
Dept of = Neurosciences
University Medical Center = (CMU)
1 Michel-Servet - 1211 GENEVA - = CH

Tel 0041 (0) 22 379 5324
Fax 0041 (0) 22 379 5402
=

Dave Flitney, = IT Manager
Oxford Centre for Functional MRI = of the Brain
E:[log in to unmask] = W:+44-1865-222713 F:+44-1865-222717

=

= --Apple-Mail-1--591614522-- ========================================================================= Date: Mon, 15 Oct 2007 16:49:31 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Conrad Rockel <[log in to unmask]> Subject: TBSS - file exclusion Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Hi there, I am working through the various TBSS stages, and upon reviewing the outp= ut of tbss_3_postreg, I noticed that one of the scans is badly registered. = Is there a way I can remove this scan from the 4D image and subsequent analy= sis without having to run the previous step again (because it took several da= ys to run)? Cheers, Conrad ========================================================================= Date: Mon, 15 Oct 2007 12:48:15 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Buyean Lee <[log in to unmask]> Subject: Re: Using SIENA in monkey MRIs In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="--------MB_8C9DD6250B9B5AA_A54_92B0_FWM-M08.sysops.aol.com" ----------MB_8C9DD6250B9B5AA_A54_92B0_FWM-M08.sysops.aol.com Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii" Hi Steve, Would you let me know which files you want to see; there are three kinds of MRIs? 1. Analyze file formats in the orientation of acquisition (because the nose is too big, I couldn't use AC-PC alignment). 2. Analyze file formats after AC-PC alignment; resliced images 3. Nifti file formats after removing the neck and tissues outside the skull (these are the files works best in SIENA). I think it might be informative if I can send you the whole directory SIENA created so that you can see the report. Please let me know if you want me to do it. Thank you, Buyean -----Original Message----- From: Steve Smith <[log in to unmask]> To: [log in to unmask] Sent: Sun, 14 Oct 2007 3:35 am Subject: Re: [FSL] Using SIENA in monkey MRIs Hi,? ? Is the movement between the two timepoints after registration due to atrophy effects of interest, or "real" misregistration?? ? If you would like us to have a quick look, please upload the files in a single compressed tarfile to? http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi? and then email me the upload ID.? ? Cheers.? ? ? On 11 Oct 2007, at 03:15, Buyean Lee wrote:? ? > Hi Steve,? >? > Finally, I ran Siena with the monkey MRIs (it is just one subject); > the siena_report.html looks much better in terms of brain > extraction, coregistation, etc.? >? > This is what I did.? >? > 1. removed all tissues outside the skull and below the cerebellum.? > 2. modified the siena.sh in order to use separate BET option for > two scans.? >? > Brain extraction is much better, but coregistration by FLIRT is not > satisfactory.? > I can still see some movement in the overlap of the two coregistred > MRIs.? >? > Is there any way for me to send these two MRIs to you so that you > can take a look at them?? >? > I just don't know what else I can do to improve the BET and > coregistration further.? >? > Thank you,? >? > Buyean? >? >? >? > -----Original Message-----? > From: Steve Smith <[log in to unmask]>? > To: [log in to unmask] > Sent: Thu, 27 Sep 2007 1:20 am? > Subject: Re: [FSL] Using SIENA in monkey MRIs? >? > HI - you would also want to include a rasonable amount of the outer > skull boundary that's near the brain boundary, as this gets used in > SIENA.? >? > To choose a FOV in FSLView and then reduce the image by this:? > Click around and find the voxel coordinates (voxel not mm) that > bound the region you want to keep.? > Find the xmin and xmax values, and calculate the xsize = xmax - xmin? > Same for y and z.? >? > Then run? >? > fslroi originalheadimage reducedimage xmin xsize ymin ysize zmin zsize? >? > (do this for both timepoints separately)? >? > and feed the reducedimages into SIENA.? >? >? > Cheers.? >? > On 26 Sep 2007, at 23:52, Buyean Lee wrote:? >? > > Hi Steve,? > >? > > "Alternatively, you could work out how to reduce the field-of-> view > to just include the brain of each time point before feeding > the > output of that into SIENA. You would use FSLView on each > image to > work out the field-of-view and then use fslroi to reduce > the FOV of > the image."? > >? > > I created a whole brain mask for the animal brain using FSLView.? > >? > > But, it is not easy to understand how to use fslroi.? > >? > > fslroi ...? > >? > > I just don't know how to determine these parameters and when I am > > supposed to use the mask file in fslroi.? > >? > > Would you kindly let me know how to reduce the field-of-view to > > include the brain?? > >? > >? > > Thank you,? > >? > > Buyean? > > > ----------------------------------------------------------------------> > -----? > > Check Out the new free AIM(R) Mail -- Unlimited storage and > > industry-leading spam and email virus protection.? >? > ----------------------------------------------------------------------> -----? > Stephen M. Smith, Professor of Biomedical Engineering? > Associate Director, Oxford University FMRIB Centre? >? > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK? > +44 (0) 1865 222726 (fax 222717)? > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve? > ----------------------------------------------------------------------> -----? > Check Out the new free AIM(R) Mail -- Unlimited storage and > industry-leading spam and email virus protection.? ? ---------------------------------------------------------------------------? Stephen M. Smith, Professor of Biomedical Engineering? Associate Director, Oxford University FMRIB Centre? ? FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK? +44 (0) 1865 222726 (fax 222717)? [log in to unmask] http://www.fmrib.ox.ac.uk/~steve? ---------------------------------------------------------------------------? ________________________________________________________________________ Check Out the new free AIM(R) Mail -- Unlimited storage and industry-leading spam and email virus protection. ----------MB_8C9DD6250B9B5AA_A54_92B0_FWM-M08.sysops.aol.com Content-Transfer-Encoding: 7bit Content-Type: text/html; charset="us-ascii"
Hi Steve,

Would you let me know which files you want to see; there are three kinds of MRIs?

1. Analyze file formats in the orientation of acquisition (because the nose is too big, I couldn't use AC-PC alignment).
2. Analyze file formats after AC-PC alignment; resliced images
3. Nifti file formats after removing the neck and tissues outside the skull (these are the files works best in SIENA).

I think it might be informative if I can send you the whole directory SIENA created so that you can see the report.
Please let me know if you want me to do it.

Thank you,

Buyean

-----Original Message-----
From: Steve Smith <[log in to unmask]>
To: [log in to unmask]
Sent: Sun, 14 Oct 2007 3:35 am
Subject: Re: [FSL] Using SIENA in monkey MRIs

Hi, 
 
Is the movement between the two timepoints after registration due to atrophy effects of interest, or "real" misregistration? 
 
If you would like us to have a quick look, please upload the files in a single compressed tarfile to 
http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi 
and then email me the upload ID. 
 
Cheers. 
 
 
On 11 Oct 2007, at 03:15, Buyean Lee wrote: 
 
> Hi Steve, 

> Finally, I ran Siena with the monkey MRIs (it is just one subject); > the siena_report.html looks much better in terms of brain > extraction, coregistation, etc. 

> This is what I did. 

> 1. removed all tissues outside the skull and below the cerebellum. 
> 2. modified the siena.sh in order to use separate BET option for > two scans. 

> Brain extraction is much better, but coregistration by FLIRT is not > satisfactory. 
> I can still see some movement in the overlap of the two coregistred > MRIs. 

> Is there any way for me to send these two MRIs to you so that you > can take a look at them? 

> I just don't know what else I can do to improve the BET and > coregistration further. 

> Thank you, 

> Buyean 



> -----Original Message----- 
> From: Steve Smith <[log in to unmask]
> To: [log in to unmask] 
> Sent: Thu, 27 Sep 2007 1:20 am 
> Subject: Re: [FSL] Using SIENA in monkey MRIs 

> HI - you would also want to include a rasonable amount of the outer > skull boundary that's near the brain boundary, as this gets used in > SIENA. 

> To choose a FOV in FSLView and then reduce the image by this: 
> Click around and find the voxel coordinates (voxel not mm) that > bound the region you want to keep. 
> Find the xmin and xmax values, and calculate the xsize = xmax - xmin 
> Same for y and z. 

> Then run 

> fslroi originalheadimage reducedimage xmin xsize ymin ysize zmin zsize 

> (do this for both timepoints separately) 

> and feed the reducedimages into SIENA. 


> Cheers. 

> On 26 Sep 2007, at 23:52, Buyean Lee wrote: 

> > Hi Steve, 
> > 
> > "Alternatively, you could work out how to reduce the field-of-> view > to just include the brain of each time point before feeding > the > output of that into SIENA. You would use FSLView on each > image to > work out the field-of-view and then use fslroi to reduce > the FOV of > the image." 
> > 
> > I created a whole brain mask for the animal brain using FSLView. 
> > 
> > But, it is not easy to understand how to use fslroi. 
> > 
> > fslroi <input> <ouput> <xmin> <xsize> <ymin> ... 
> > 
> > I just don't know how to determine these parameters and when I am > > supposed to use the mask file in fslroi. 
> > 
> > Would you kindly let me know how to reduce the field-of-view to > > include the brain? 
> > 
> > 
> > Thank you, 
> > 
> > Buyean 
> > > ----------------------------------------------------------------------> > ----- 
> > Check Out the new free AIM(R) Mail -- Unlimited storage and > > industry-leading spam and email virus protection. 

> ----------------------------------------------------------------------> ----- 
> Stephen M. Smith, Professor of Biomedical Engineering 
> Associate Director, Oxford University FMRIB Centre 

> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK 
> +44 (0) 1865 222726 (fax 222717) 
> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve 
> ----------------------------------------------------------------------> ----- 
> Check Out the new free AIM(R) Mail -- Unlimited storage and > industry-leading spam and email virus protection. 
 
--------------------------------------------------------------------------- 
Stephen M. Smith, Professor of Biomedical Engineering 
Associate Director, Oxford University FMRIB Centre 
 
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK 
+44 (0) 1865 222726 (fax 222717) 
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve 
--------------------------------------------------------------------------- 

Check Out the new free AIM(R) Mail -- Unlimited storage and industry-leading spam and email virus protection.
----------MB_8C9DD6250B9B5AA_A54_92B0_FWM-M08.sysops.aol.com-- ========================================================================= Date: Mon, 15 Oct 2007 18:03:38 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Patrick Hales <[log in to unmask]> Subject: Re: Problem with DTIFit In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit It worked! DTIFit will now process the images. Thanks very much. One last thing - just to clarify, a '1' in the bvecs file means sqrt (2)? Why isn't it 1/sqrt(2)? When I produced these images I used factors of 1/sqrt(2) in each direction, so the total magnitude would always equal 1. For instance, in my 2nd volume (1 1 0) I multiplied the total gradient by (1/sqrt(2)) in the x direction, and the same in the y direction. Sorry if this is a stupid question, but how do I tell FSL to use 1/sqrt(2) as a unit length? Thanks Patrick On 15 Oct 2007, at 16:13, Tim Behrens wrote: > PS - > >> 0 1 0 1 0 1 -1 >> 0 1 1 0 1 -1 0 >> 0 0 1 1 -1 0 1 >> >> i.e. the gradients directions have been normalized to unit length, > > > These are not unit length. They have a length of sqrt(2) > > This does not matter for the new version of FDT, but probably does > matter for the old version you are working with > > T > > > > > On 15 Oct 2007, at 15:53, Patrick Hales wrote: > >> Hi >> >> Thanks again for the help. I've tried running bedpost_datacheck on >> my data (after renaming everything to the appropriate filenames). >> The output I get is: >> >> bedpost_data/data >> data_type INT16 >> dim1 128 >> dim2 128 >> dim3 7 >> dim4 1 >> datatype 4 >> pixdim1 0.2343750000 >> pixdim2 0.2343750000 >> pixdim3 1.0000000000 >> pixdim4 1.0000000000 >> cal_max 25500.0000 >> cal_min 40.0000 >> file_type NIFTI-1+ >> >> bedpost_data/nodif does not exist >> bedpost_data/nodif_brain_mask >> data_type INT16 >> dim1 128 >> dim2 128 >> dim3 7 >> dim4 1 >> datatype 4 >> pixdim1 0.2343750000 >> pixdim2 0.2343750000 >> pixdim3 1.0000000000 >> pixdim4 1.0000000000 >> cal_max 4.0000 >> cal_min 0.0000 >> file_type NIFTI-1+ >> >> num lines in bedpost_data/bvals >> 1 >> num words in bedpost_data/bvals >> 7 >> num lines in bedpost_data/bvecs >> 3 >> num words in bedpost_data/bvecs >> 21 >> number of elements in bvals is not equal to number of vols in data >> number of elements per line in bvecs is not equal to number of >> vols in data >> >> I don't understand what's wrong with my bvals and bvecs files. My >> 'data' file should now be made up of 7 volumes, the 1st with b=0 >> and the other 6 with diffusion weighting in various directions. My >> bvals file looks like this: >> >> 0 1025 1048 826 1003 582 826 >> >> and my bvecs file looks like this: >> >> 0 1 0 1 0 1 -1 >> 0 1 1 0 1 -1 0 >> 0 0 1 1 -1 0 1 >> >> i.e. the gradients directions have been normalized to unit length, >> and the vectors run in columns. Could you tell me what I'm doing >> wrong with these files? - I've read the help and it sounds like I >> have set them up correctly, but I'm obviously doing something wrong. >> >> Thanks very much >> >> Patrick >> >> >> >> >> >> >> On 15 Oct 2007, at 14:28, Tim Behrens wrote: >> >>> Although - if you have the old version of FSL, it will be called >>> bedpost_datacheck! >>> T >>> On 15 Oct 2007, at 13:00, Matt Glasser wrote: >>> >>>> Hi, >>>> >>>> Fslmerge is in the latest version of FSL whereas avwmerge is in >>>> earlier >>>> versions. You could navigate to the data directory with the >>>> commandline and >>>> try running bedpostx_datacheck and post that output here. >>>> >>>> Peace, >>>> >>>> Matt. >>>> >>>> -----Original Message----- >>>> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] >>>> On Behalf >>>> Of Patrick Hales >>>> Sent: Monday, October 15, 2007 6:36 AM >>>> To: [log in to unmask] >>>> Subject: Re: [FSL] Problem with DTIFit >>>> >>>> Hi >>>> >>>> Thanks for that. The command 'fslmerge' isn't recognized for some >>>> reason, however, I have been able to make my 4D data file using >>>> 'avwmerge' instead. Unfortunately, when I try to process the >>>> data in >>>> DTIFit, I now get the error message: "Error: child killed: bus >>>> error". Would you be able to tell me how to fix this? >>>> >>>> Thanks very much for your help >>>> >>>> Patrick >>>> >>>> >>>> >>>> >>>> On 15 Oct 2007, at 11:14, Tim Behrens wrote: >>>> >>>>> You need to merge your data from different diffusion directions >>>>> into a single 4D nifti file - you can use fslmerge to do this. >>>>> >>>>> T >>>>> On 15 Oct 2007, at 09:52, Patrick Hales wrote: >>>>> >>>>>> Hi >>>>>> >>>>>> This is the first time I have tried using FSL, and I am trying to >>>>>> get the DTIFit program to work. I have >>>>>> a series of diffusion weighted images, each of which are >>>>>> stored as >>>>>> individual fid files (created using >>>>>> Varian's VnmrJ software). There are seven images in total, one >>>>>> with b=0 and 6 with diffusion >>>>>> gradients applied in non co-linear directions. I have converted >>>>>> each of these into Nifti format, and >>>>>> stored them all in a directory. I used FSLview to create my >>>>>> binary >>>>>> mask, and have created text files >>>>>> with my gradient directions (all normalized to unit length), and >>>>>> my b-values (in units of s/mm2). In >>>>>> the DTIFit program, I have selected 'specify input files >>>>>> manually', and supplied it with the directory >>>>>> containing my nifti files, and the filenames of my gradients >>>>>> and b- >>>>>> values text files. >>>>>> >>>>>> When I press 'go', I get a number of error messages, such as: >>>>>> 'ERROR: nifti_image_read(....directory >>>>>> name....)can't open header file', 'ERROR: nifti_image_open >>>>>> (...):bad header info', 'Error: failed to open >>>>>> file', etc. I get the same errors when I try it with analyze >>>>>> files. >>>>>> >>>>>> Would you be able to tell me what I'm doing wrong? Thanks very >>>>>> much >>>>>> >>>>>> Patrick Hales ========================================================================= Date: Mon, 15 Oct 2007 20:49:41 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: =?iso-8859-1?q?Gwena=EBlle=20DOUAUD?= <[log in to unmask]> Subject: RE : [FSL] TBSS - file exclusion In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Hi Conrad, I'm not sure if somebody has already answered your question, it seems that I receive all the emails quite delayed today. Anyway, sure, if you can not fix it, you can always remove this image from your 4D dataset. If the bad scan is the volume #i in _fslview_ among your n images, you just need to use fslroi with something like: fslroi all_FA all_FA_1 0 i fslroi all_FA all_FA_2 i+1 n-(i+1) fslmerge -t all_FA all_FA_1 all_FA_2 Also, please just make sure that the badly registered scan has not mislead the mean_FA and mean_FA_skeleton creation (hopefully not). Hope this helps, Gwenaelle --- Conrad Rockel <[log in to unmask]> a =E9crit : > Hi there, >=20 > I am working through the various TBSS stages, and > upon reviewing the output > of tbss_3_postreg, I noticed that one of the scans > is badly registered. Is > there a way I can remove this scan from the 4D image > and subsequent analysis > without having to run the previous step again > (because it took several days > to run)? >=20 > Cheers, >=20 > Conrad >=20 ___________________________________________________________________= __________=20 Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! M= ail=20 ========================================================================= Date: Mon, 15 Oct 2007 20:20:27 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Markus Gschwind <[log in to unmask]> Subject: Re: DTI processing In-Reply-To: <[log in to unmask]> MIME-version: 1.0 Content-transfer-encoding: 7BIT Content-disposition: inline Content-type: text/plain; charset=ISO-8859-1; DelSp=Yes; format=flowed Hi Dave! Thanks a lot! That was it: > try thresholding it appropriately > (i.e., to isolate regions of similar FA) then setting it to Mask/Label > in the image info dialog then switch to 3D. The problem was, that I had FIRST to adjust threshold and mask/label AND THEN switch to 3D. Otherwise my fslview would get stuck (maybe due tu costs of the vm-machine?). Very nice result! Markus > >>>> 1. >>>> I can open FA data in FSLveiw 2D nicely, but I cannot open any >>>> FA-data (produced in FSL3 on Cygwin) in FSLview 3D. That means: >>>> It atually works but I just don't see anything! Is that normal? >>> >>> Have you tried setting the FSLView display range to 0:1? >> >> Yes exactly. But I dont see anything. Acutally in the terminal it tells me: >> >> "ERROR: In /home/duncan/VTK/Graphics/vtkPolyDataNormals.cxx, line 94 >> vtkPolyDataNormals (0xa2fccb8): No data to generate normals for!" > > Are you talking about 3D view or all views? Can you see your FA data > in 2D view? > > There is no specific mode for showing FA data in 3D. 3D view will do a > surface for the main image - threshold it with the box at the top of > the display. It will also: blend in stats images and show an iso- > surface for them; show the surface of mask images. If you want to view > DTI data then perhaps you could try thresholding it appropriately > (i.e., to isolate regions of similar FA) then setting it to Mask/Label > in the image info dialog then switch to 3D. > >> Is that a compatibility problem of FSL 3 and FSL 4? >> >> Thanks again, cheers, >> Markus >> >> >>>> -- >>>> >>>> Markus Gschwind, M.D. >>>> Laboratory for Neurology and Imaging of Cognition >>>> Dept of Neurosciences >>>> University Medical Center (CMU) >>>> 1 Michel-Servet - 1211 GENEVA - CH >>>> >>>> Tel 0041 (0) 22 379 5324 >>>> Fax 0041 (0) 22 379 5402 >>>> email: [log in to unmask] >>>> http://labnic.unige.ch >>> >>> >>> --------------------------------------------------------------------- >>> ------ >>> Stephen M. Smith, Professor of Biomedical Engineering >>> Associate Director, Oxford University FMRIB Centre >>> >>> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >>> +44 (0) 1865 222726 (fax 222717) >>> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >>> --------------------------------------------------------------------- >>> ------ >> >> >> >> -- >> >> Markus Gschwind, M.D. >> Laboratory for Neurology and Imaging of Cognition >> Dept of Neurosciences >> University Medical Center (CMU) >> 1 Michel-Servet - 1211 GENEVA - CH >> >> Tel 0041 (0) 22 379 5324 >> Fax 0041 (0) 22 379 5402 >> email: [log in to unmask] >> http://labnic.unige.ch > > -- > Cheers, Dave > > Dave Flitney, IT Manager > Oxford Centre for Functional MRI of the Brain > E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 > URL: http://www.fmrib.ox.ac.uk/~flitney -- Markus Gschwind, M.D. Laboratory for Neurology and Imaging of Cognition Dept of Neurosciences University Medical Center (CMU) 1 Michel-Servet - 1211 GENEVA - CH Tel 0041 (0) 22 379 5324 Fax 0041 (0) 22 379 5402 email: [log in to unmask] http://labnic.unige.ch ========================================================================= Date: Mon, 15 Oct 2007 15:55:47 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matt Glasser <[log in to unmask]> Organization: ma-tea.com Subject: Re: DTI processing In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit That bug exists in native installs also. Peace, Matt. -----Original Message----- From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Markus Gschwind Sent: Monday, October 15, 2007 2:20 PM To: [log in to unmask] Subject: Re: [FSL] DTI processing Hi Dave! Thanks a lot! That was it: > try thresholding it appropriately > (i.e., to isolate regions of similar FA) then setting it to Mask/Label > in the image info dialog then switch to 3D. The problem was, that I had FIRST to adjust threshold and mask/label AND THEN switch to 3D. Otherwise my fslview would get stuck (maybe due tu costs of the vm-machine?). Very nice result! Markus > >>>> 1. >>>> I can open FA data in FSLveiw 2D nicely, but I cannot open any >>>> FA-data (produced in FSL3 on Cygwin) in FSLview 3D. That means: >>>> It atually works but I just don't see anything! Is that normal? >>> >>> Have you tried setting the FSLView display range to 0:1? >> >> Yes exactly. But I dont see anything. Acutally in the terminal it tells me: >> >> "ERROR: In /home/duncan/VTK/Graphics/vtkPolyDataNormals.cxx, line 94 >> vtkPolyDataNormals (0xa2fccb8): No data to generate normals for!" > > Are you talking about 3D view or all views? Can you see your FA data > in 2D view? > > There is no specific mode for showing FA data in 3D. 3D view will do a > surface for the main image - threshold it with the box at the top of > the display. It will also: blend in stats images and show an iso- > surface for them; show the surface of mask images. If you want to view > DTI data then perhaps you could try thresholding it appropriately > (i.e., to isolate regions of similar FA) then setting it to Mask/Label > in the image info dialog then switch to 3D. > >> Is that a compatibility problem of FSL 3 and FSL 4? >> >> Thanks again, cheers, >> Markus >> >> >>>> -- >>>> >>>> Markus Gschwind, M.D. >>>> Laboratory for Neurology and Imaging of Cognition >>>> Dept of Neurosciences >>>> University Medical Center (CMU) >>>> 1 Michel-Servet - 1211 GENEVA - CH >>>> >>>> Tel 0041 (0) 22 379 5324 >>>> Fax 0041 (0) 22 379 5402 >>>> email: [log in to unmask] >>>> http://labnic.unige.ch >>> >>> >>> --------------------------------------------------------------------- >>> ------ >>> Stephen M. Smith, Professor of Biomedical Engineering >>> Associate Director, Oxford University FMRIB Centre >>> >>> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >>> +44 (0) 1865 222726 (fax 222717) >>> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >>> --------------------------------------------------------------------- >>> ------ >> >> >> >> -- >> >> Markus Gschwind, M.D. >> Laboratory for Neurology and Imaging of Cognition >> Dept of Neurosciences >> University Medical Center (CMU) >> 1 Michel-Servet - 1211 GENEVA - CH >> >> Tel 0041 (0) 22 379 5324 >> Fax 0041 (0) 22 379 5402 >> email: [log in to unmask] >> http://labnic.unige.ch > > -- > Cheers, Dave > > Dave Flitney, IT Manager > Oxford Centre for Functional MRI of the Brain > E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717 > URL: http://www.fmrib.ox.ac.uk/~flitney -- Markus Gschwind, M.D. Laboratory for Neurology and Imaging of Cognition Dept of Neurosciences University Medical Center (CMU) 1 Michel-Servet - 1211 GENEVA - CH Tel 0041 (0) 22 379 5324 Fax 0041 (0) 22 379 5402 email: [log in to unmask] http://labnic.unige.ch ========================================================================= Date: Mon, 15 Oct 2007 22:20:59 +0200 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Volkmar Glauche <[log in to unmask]> Subject: Re: Problem with DTIFit In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; DelSp="Yes"; format="flowed" Content-Disposition: inline Content-Transfer-Encoding: quoted-printable Dear Patrick, the length of a 3-vector is defined as sqrt(x^2 + y^2 + z^2) Thus, a vector (1 -1 0) has a length of sqrt(1+1+0) =3D sqrt(2). Volkmar Zitat von Patrick Hales <[log in to unmask]>: > It worked! DTIFit will now process the images. Thanks very much. > > One last thing - just to clarify, a '1' in the bvecs file means > sqrt(2)? Why isn't it 1/sqrt(2)? When I produced these images I used > factors of 1/sqrt(2) in each direction, so the total magnitude would > always equal 1. For instance, in my 2nd volume (1 1 0) I multiplied the > total gradient by (1/sqrt(2)) in the x direction, and the same in the y > direction. Sorry if this is a stupid question, but how do I tell FSL to > use 1/sqrt(2) as a unit length? > > Thanks > > Patrick > > > > > On 15 Oct 2007, at 16:13, Tim Behrens wrote: > >> PS - >> >>> 0 1 0 1 0 1 -1 >>> 0 1 1 0 1 -1 0 >>> 0 0 1 1 -1 0 1 >>> >>> i.e. the gradients directions have been normalized to unit length, >> >> >> These are not unit length. They have a length of sqrt(2) >> >> This does not matter for the new version of FDT, but probably does =20 >> matter for the old version you are working with >> >> T >> >> >> >> >> On 15 Oct 2007, at 15:53, Patrick Hales wrote: >> >>> Hi >>> >>> Thanks again for the help. I've tried running bedpost_datacheck on =20 >>> my data (after renaming everything to the appropriate filenames). =20 >>> The output I get is: >>> >>> bedpost_data/data >>> data_type INT16 >>> dim1 128 >>> dim2 128 >>> dim3 7 >>> dim4 1 >>> datatype 4 >>> pixdim1 0.2343750000 >>> pixdim2 0.2343750000 >>> pixdim3 1.0000000000 >>> pixdim4 1.0000000000 >>> cal_max 25500.0000 >>> cal_min 40.0000 >>> file_type NIFTI-1+ >>> >>> bedpost_data/nodif does not exist >>> bedpost_data/nodif_brain_mask >>> data_type INT16 >>> dim1 128 >>> dim2 128 >>> dim3 7 >>> dim4 1 >>> datatype 4 >>> pixdim1 0.2343750000 >>> pixdim2 0.2343750000 >>> pixdim3 1.0000000000 >>> pixdim4 1.0000000000 >>> cal_max 4.0000 >>> cal_min 0.0000 >>> file_type NIFTI-1+ >>> >>> num lines in bedpost_data/bvals >>> 1 >>> num words in bedpost_data/bvals >>> 7 >>> num lines in bedpost_data/bvecs >>> 3 >>> num words in bedpost_data/bvecs >>> 21 >>> number of elements in bvals is not equal to number of vols in data >>> number of elements per line in bvecs is not equal to number of vols in d= ata >>> >>> I don't understand what's wrong with my bvals and bvecs files. My =20 >>> 'data' file should now be made up of 7 volumes, the 1st with b=3D0 =20 >>> and the other 6 with diffusion weighting in various directions. My =20 >>> bvals file looks like this: >>> >>> 0 1025 1048 826 1003 582 826 >>> >>> and my bvecs file looks like this: >>> >>> 0 1 0 1 0 1 -1 >>> 0 1 1 0 1 -1 0 >>> 0 0 1 1 -1 0 1 >>> >>> i.e. the gradients directions have been normalized to unit length, =20 >>> and the vectors run in columns. Could you tell me what I'm doing =20 >>> wrong with these files? - I've read the help and it sounds like I =20 >>> have set them up correctly, but I'm obviously doing something wrong. >>> >>> Thanks very much >>> >>> Patrick >>> >>> >>> >>> >>> >>> >>> On 15 Oct 2007, at 14:28, Tim Behrens wrote: >>> >>>> Although - if you have the old version of FSL, it will be called =20 >>>> bedpost_datacheck! >>>> T >>>> On 15 Oct 2007, at 13:00, Matt Glasser wrote: >>>> >>>>> Hi, >>>>> >>>>> Fslmerge is in the latest version of FSL whereas avwmerge is in earlie= r >>>>> versions. You could navigate to the data directory with the =20 >>>>> commandline and >>>>> try running bedpostx_datacheck and post that output here. >>>>> >>>>> Peace, >>>>> >>>>> Matt. >>>>> >>>>> -----Original Message----- >>>>> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] =20 >>>>> On Behalf >>>>> Of Patrick Hales >>>>> Sent: Monday, October 15, 2007 6:36 AM >>>>> To: [log in to unmask] >>>>> Subject: Re: [FSL] Problem with DTIFit >>>>> >>>>> Hi >>>>> >>>>> Thanks for that. The command 'fslmerge' isn't recognized for some >>>>> reason, however, I have been able to make my 4D data file using >>>>> 'avwmerge' instead. Unfortunately, when I try to process the data in >>>>> DTIFit, I now get the error message: "Error: child killed: bus >>>>> error". Would you be able to tell me how to fix this? >>>>> >>>>> Thanks very much for your help >>>>> >>>>> Patrick >>>>> >>>>> >>>>> >>>>> >>>>> On 15 Oct 2007, at 11:14, Tim Behrens wrote: >>>>> >>>>>> You need to merge your data from different diffusion directions >>>>>> into a single 4D nifti file - you can use fslmerge to do this. >>>>>> >>>>>> T >>>>>> On 15 Oct 2007, at 09:52, Patrick Hales wrote: >>>>>> >>>>>>> Hi >>>>>>> >>>>>>> This is the first time I have tried using FSL, and I am trying to >>>>>>> get the DTIFit program to work. I have >>>>>>> a series of diffusion weighted images, each of which are stored as >>>>>>> individual fid files (created using >>>>>>> Varian's VnmrJ software). There are seven images in total, one >>>>>>> with b=3D0 and 6 with diffusion >>>>>>> gradients applied in non co-linear directions. I have converted >>>>>>> each of these into Nifti format, and >>>>>>> stored them all in a directory. I used FSLview to create my binary >>>>>>> mask, and have created text files >>>>>>> with my gradient directions (all normalized to unit length), and >>>>>>> my b-values (in units of s/mm2). In >>>>>>> the DTIFit program, I have selected 'specify input files >>>>>>> manually', and supplied it with the directory >>>>>>> containing my nifti files, and the filenames of my gradients and b- >>>>>>> values text files. >>>>>>> >>>>>>> When I press 'go', I get a number of error messages, such as: >>>>>>> 'ERROR: nifti_image_read(....directory >>>>>>> name....)can't open header file', 'ERROR: nifti_image_open >>>>>>> (...):bad header info', 'Error: failed to open >>>>>>> file', etc. I get the same errors when I try it with analyze files. >>>>>>> >>>>>>> Would you be able to tell me what I'm doing wrong? Thanks very much >>>>>>> >>>>>>> Patrick Hales ========================================================================= Date: Mon, 15 Oct 2007 16:25:05 -0500 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Michelle Voss <[log in to unmask]> Subject: group melodic and individual IC effect sizes MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_39393_14684063.1192483505372" ------=_Part_39393_14684063.1192483505372 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline hello, i've run a group melodic analysis, and i want to correlate the extent to which subjects exhibit a component (say the RSN) with another variable (say age). this info is captured in the subjects/session mode plot for individuals about the group mean, and in the box plot to show formally the group variation about the mean--in that-- the extent to which a component is represented and regions in it functionally covary for an individual is represented by their individual effect size and for a group by mean effect size. when you contrast two groups (say male and female) within the group PICA analysis, this contrasts the difference in mean effect (or representation of that group) for a component. is my understanding of the melodic report right? i.e., the individual effects sizes in the subjects/session mode are meaningful in looking for individual differences in exemplifying a component (?) thanks so much, Michelle ------=_Part_39393_14684063.1192483505372 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline hello,

i've run a group melodic analysis, and i want to correlate the extent to which subjects exhibit a component (say the RSN) with another variable (say age).  this info is captured in the subjects/session mode plot for individuals about the group mean, and in the box plot to show formally the group variation about the mean--in that-- the extent to which a component is represented and regions in it functionally covary for an individual is represented by their individual effect size and for a group by mean effect size.  when you contrast two groups (say male and female) within the group PICA analysis, this contrasts the difference in mean effect (or representation of that group) for a component. 

is my understanding of the melodic report right?  i.e., the individual effects sizes in the subjects/session mode are meaningful in looking for individual differences in exemplifying a component (?)

thanks so much,

Michelle

 


------=_Part_39393_14684063.1192483505372-- ========================================================================= Date: Mon, 15 Oct 2007 23:01:44 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Kimberly Carpenter <[log in to unmask]> Subject: WARNING::in calculating COG, total = 0.0 Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="ISO-8859-1" Hello, I am having problems running 1st level analyses on a few of my subject= s. During the prestats stage of the analysis I get a repeated error of=20 WARNING::in calculating COG, total =3D 0.0 I see that previous posts have said that "The COG total=3D0 warning occu= rs when the images are empty - that is, all zeros." However, when I look at= my raw data using a Matlab script I have images. Does anyone know what may = be causing this error? Thanks! ========================================================================= Date: Mon, 15 Oct 2007 23:18:50 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Tara Phillips <[log in to unmask]> Subject: Fedora core 7 and FSL Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="UTF-8" Hello, I am trying to run fsl in fedora core 7 on a 32-bit linus. The fsl menu appears and all of the options are functioning except for 'FSL view'. Wh= en I click on FSL view, the following error appears: libqt-mt.so.3: cannot open shared object file: No such file or directory Indeed, I do not have this library. I tried to download it but was sent = on a goose-chase. Do you have any suggestions for resolving this? Regards, Tara ========================================================================= Date: Mon, 15 Oct 2007 18:50:12 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: "Zhang, Xiaochu (NIH/NIDA) [F]" <[log in to unmask]> Subject: Re: RE : [FSL] The orientation problem in VBM analysis In-Reply-To: A<[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: base64 VGhhbmsgeW91IHZlcnkgbXVjaCBmb3IgeW91ciBoZWxwISBJIHJlYWQgdGhlIG1haWwgeW91IGdh dmUgdG8gbWUuIEhvd2V2ZXIsIEkgZm91bmQgd2UgaGF2ZSBkaWZmZXJlbnQgcHJvYmxlbXMuIElu IG91ciBjYXNlLCB3ZSBkb27igJl0IHdhbnQgdG8gb25seSBjaGFuZ2UgTC9SLiBIb3dldmVyLCBX ZSB3YW50IHRvIGNoYW5nZSB0aGUgeSBheGlhbCBhbmQgdGhlIHogYXhpYWwuIE5vdywgdGhlIHog KHkpIGF4aWFsIHNob3VsZCBiZSBBL1AgKFMvSSkgYnV0IGlzIFMvSSAoQS9QKSBub3cuIEkgY2hh bmdlZCB0aGUgb3JpZW50YXRpb24gaW4gTklGVEkgdmlhICJmc2xvcmllbnQiLiBIb3dldmVyLCBh ZnRlciAidGJzc18xX3ByZXByb2Nlc3MiLCBpdCBpcyB3cm9uZywgYWdhaW4uIFRoZSAiZnNsb3Jp ZW50IiBsb29rcyBzZWVtbHkgbm90IHdvcmsuDQpUaGUgImZzbHN3YXBkaW0iIGlzIHdvcmtlZC4g SG93ZXZlciwgSSBkb24ndCBrbm93IGhvdyB0byBjaGFuZ2UgeSBheGlhbCBhbmQgeiBheGlhbCBi eSBpdC4NCkNvdWxkIHlvdSBoZWxwIG1lPyBJJ20gZWFnZXIgdG8geW91ciBoZWxwLg0KVGhhbmtz LCBhZ2FpbiENClhpYW9jaHUNCg0KLS0tLS1PcmlnaW5hbCBNZXNzYWdlLS0tLS0NCkZyb206IEd3 ZW5hw6tsbGUgRE9VQVVEIFttYWlsdG86Z3dlbmFlbGxlZG91YXVkQFlBSE9PLkZSXSANClNlbnQ6 IDIwMDflubQxMOaciDEz5pelIDIyOjA1DQpUbzogRlNMQEpJU0NNQUlMLkFDLlVLDQpTdWJqZWN0 OiBbRlNMXSBSRSA6IFtGU0xdIFRoZSBvcmllbnRhdGlvbiBwcm9ibGVtIGluIFZCTSBhbmFseXNp cw0KDQpIaSBhZ2FpbiBYaWFvY2h1LA0KDQpmb3IgdGhpcyBvcmllbnRhdGlvbmFsIHByb2JsZW0s IHBsZWFzZSBsb29rIGF0DQpodHRwOi8vd3d3Lmppc2NtYWlsLmFjLnVrL2NnaS1iaW4vd2ViYWRt aW4/QTI9aW5kMDcwOSZMPUZTTCZQPVI0NDI0OCZJPS0zDQpvbiB0aGUgc2FtZSBzdWJqZWN0Li4u DQoNCkNoZWVycywNCkd3ZW5hw6tsbGUNCg0KLS0tICJaaGFuZywgWGlhb2NodSAoTklIL05JREEp IFtGXSINCjx6aGFuZ3gyQElOVFJBLk5JREEuTklILkdPVj4gYSDDqWNyaXQgOg0KDQo+IEhpIGV4 cGVydHMsDQo+IA0KPiBJIGFtIHdvcmtpbmcgb24gdGhlIFZCTSBhbmFseXNpcy4gRm9yIHRoZSBk YXRhIGlzDQo+IGZyb20gbXVsdGlwbGUgc3R1ZGllcywNCj4gc29tZSBpcyBzYWdpdGlhbCBhbmQg c29tZSBpcyBheGlhbC4gQWZ0ZXIgInRic3NfMSINCj4gYW5kIGNoZWNrIHRoZQ0KPiBzbGljZWRp ciwgdGhlIGRhdGEgc2hvd2VkIGluY29uc2lzdGVudCBvcmllbnRhdGlvbi4NCj4gV2hvIGNhbiBo ZWxwIG1lIGFuZA0KPiB0ZWxsIG1lIGhvdyB0byBjaGFuZ2UgdGhlIG9yaWVudGF0aW9uPw0KPiAN Cj4gVGhhbmtzIGEgbG90IQ0KPiANCj4gIA0KPiANCj4gWGlhb2NodSBaaGFuZyBQaC5EDQo+IA0K PiBWaXNpdGluZyBSZXNlYXJjaCBGZWxsb3cNCj4gDQo+IE5JSC9OSURBLUlSUA0KPiANCj4gNTUw MCBOYXRoYW4gU2hvY2sgRHJpdmUNCj4gDQo+IEJhbHRpbW9yZSBNRCAyMTIyNA0KPiANCj4gIA0K PiANCj4gVGVsOiAoNDEwKSAtIDU1MCAtIDE0NDAgZXh0LiA0MzQNCj4gDQo+ICANCj4gDQo+IA0K DQoNCg0KICAgICAgX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19f X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18gDQpOZSBnYXJkZXogcGx1cyBxdSd1bmUg c2V1bGUgYWRyZXNzZSBtYWlsICEgQ29waWV6IHZvcyBtYWlscyB2ZXJzIFlhaG9vISBNYWlsIA0K ========================================================================= Date: Mon, 15 Oct 2007 19:17:54 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matt Glasser <[log in to unmask]> Subject: MNI Space in MRIcro and FSL MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0028_01C80F60.16BAD840" This is a multi-part message in MIME format. ------=_NextPart_000_0028_01C80F60.16BAD840 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Hi, I have a set of ROIs from SPM in MNI space that I want to load into FSL. When I take the standard FSL brain MNI152_T1_1mm_brain and load it in MRIcro and overlay the ROI, it looks fine, like the first screenshot: http://www.ma-tea.com/~Matt/MRIcroOnMNI.png However, if I resample the 4X4X4mm ROI so it can be overlayed on this brain in FSLView, using: flirt -in 4X4X4mmROI -ref MNI152_T1_1mm_brain -applyxfm -init eye.mat -out 1X1X1mmROI where eye.mat is the identity matrix: 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 The result is incorrect, shown by the second screenshot in FSLView: http://www.ma-tea.com/~Matt/FSLOnMNI.png The ROI file header is the following: I would appreciate any assistance with this as I am unfamiliar with this type of issue if it has to do with the analyze file format or header. Thanks, Matt. ------=_NextPart_000_0028_01C80F60.16BAD840 Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable

Hi,

 

I have a set of ROIs from SPM in MNI space that I = want to load into FSL.  When I take the standard FSL brain = MNI152_T1_1mm_brain and load it in MRIcro and overlay the ROI, it looks fine, like the first = screenshot: http://www.ma-tea.co= m/~Matt/MRIcroOnMNI.png  However, if I resample the 4X4X4mm ROI so it can be overlayed on this brain in FSLView, using:

 

flirt –in 4X4X4mmROI –ref = MNI152_T1_1mm_brain –applyxfm –init eye.mat –out 1X1X1mmROI where eye.mat = is the identity matrix:

 

1 0 0 0

0 1 0 0

0 0 1 0

0 0 0 1

 

The result is incorrect, shown by the second = screenshot in FSLView: http://www.ma-tea.com/~= Matt/FSLOnMNI.png

 

The ROI file header is the = following:

 

<nifti_image

  nifti_type =3D = 'ANALYZE-7.5'

  image_offset =3D = '0'

  ndim =3D '4'

  nx =3D '41'

  ny =3D '48'

  nz =3D '35'

  nt =3D '1'

  dx =3D '4'

  dy =3D '4'

  dz =3D '4'

  dt =3D '1'

  datatype =3D '16'

  nbyper =3D '4'

  byteorder =3D = 'LSB_FIRST'

  cal_min =3D '0'

  cal_max =3D '1'

  descrip =3D = 'FSL4.0'

  sform_code =3D = '2'

  sto_xyz_matrix =3D '-4 0 0 80 0 4 0 -112 0 0 4 = -52 0 0 0 1'

  num_ext =3D '0'

/>

 

I would appreciate any assistance with this as I am unfamiliar with this type of issue if it has to do with the analyze file = format or header.

 

Thanks,

 

Matt.

------=_NextPart_000_0028_01C80F60.16BAD840-- ========================================================================= Date: Tue, 16 Oct 2007 04:41:42 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Michael Milham <[log in to unmask]> Subject: Post-Doctoral Position Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain COGNITIVE NEUROSCIENCE POST-DOCTORAL POSITION=20 Phyllis Green and Randolph C?wen Institute for Pediatric Neuroscience, NY= U Child Study Center,=20 New York University School of Medicine, Department of Child and Adolescen= t Psychiatry The Institute for Pediatric Neuroscience is currently recruiting a post-d= octoral fellow interested in=20 identifying and/or characterizing abnormalities in the neural substrates = of fundamental cognitive=20 processes, as related to psychiatric disorders commonly encountered in ch= ildren and adolescents=20 (e.g., ADHD, autism, pediatric anxiety disorders, Tourette's).=20=20 The fellow will work as part of a multidisciplinary team, integrating the= findings of a broad=20 spectrum of approaches including: -=09resting state functional MRI -=09task-based functional MRI -=09diffusion tensor imaging -=09voxel-based morphometry -=09imaging studies in non-human primates -=09EEG/ERP=92s Qualifications: The minimum qualifications for a successful candidate inc= lude:=20 - completed PhD in cognitive neuroscience, cognitive psychology, or neuro= science - significant prior neuroimaging experience with functional MRI (prior ex= perience with EEG is a=20 plus, but not required) - strong skills in usage of one or more common functional neuroimaging (F= SL, SPM, AFNI)=20 packages=20=20 - programming experience in Matlab, C/C++ or similar platform=20 Anticipated start date: July 1, 2008.=20 How to apply: This is an electronic only application process. Please em= ail a cover letter,=20 curriculum vitae, and the names of 3 references to Michael Milham, M.D., = Ph.D.,=20 [log in to unmask], and F. Xavier Castellanos, M.D., Francisco.Castella= [log in to unmask] Feel free to contact us with any questions prior to applying for the posi= tion.=20 The New York University School of Medicine was founded in 1841 and is an = equal opportunity=20 affirmative action employer. Women and minority candidates are encourage= d to apply. ========================================================================= Date: Mon, 15 Oct 2007 21:33:18 -0700 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Jeanette Mumford <[log in to unmask]> Subject: recreating varcopes MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi, I was just trying to recreate the varcope image from a first level analysis using the sigmasquareds and threshac1 images as well as the design.mat file. My result is close, but I'm wondering why it isn't closer. Here's what I've done 1. Create the correlation matrix V using the first element of threshac1 as the main diagonal, 2nd element as the first diagonal above and below the main diagonal and so forth. 2. Demean the design matrix that's stored in design.mat (call it X) 3. Calculate inv(X'*inv(V)*X)*sigma^2 Where sigma^2 is the sigmasquareds value for that voxel and take the appropriate contrast of this matrix to correspond with my varcope of interest. My method seems to produce a variance that is consistently smaller than the varcope, but within 20% or so. Did I miss something? Highpass filtering was used in the analysis (both the data and the design), but I figured that would already be incorporated in all the pieces that I used. The analysis was originally carried out in Feat version 5.43. Thanks! Jeanette ========================================================================= Date: Tue, 16 Oct 2007 00:38:17 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Marc Dubin <[log in to unmask]> Subject: Re: mean_FA_skeleton and brain misalignment Comments: cc: [log in to unmask], [log in to unmask] In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_25261_26458503.1192509497562" ------=_Part_25261_26458503.1192509497562 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi, The overlay of the individual FA files with FMRIB58_FA seems to be good but there may be a problem with all_FA - there is a large part of the frontal lobe bilaterally that got cut out, I think during the nonlinear transformation to FMRIB58_FA. The 31 files for the 31 individual subjects are included as well as all_FA in 542195. Best, Marc On 10/15/07, Steve Smith <[log in to unmask]> wrote: > > Hi - I'm suprised that you get this with the -T option. If you load > in all_FA first into FSLView, and then overlay the FMRIB58_FA > (threshold this, change its colour to red-yellow, and make it a > little transparent) and then move through the timepoints, do all > subjects behave in a similar way to what you've described? > > We can have a look and advise on what's happening; > Please upload the files in a single compressed tarfile to > http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi > and then email me the upload ID. > > Cheers. > > > > On 15 Oct 2007, at 02:36, Marc Dubin wrote: > > > Hello FSL Users, > > > > Using the TBSS tools I have been carrying out the tbss_2_reg step > > with the -T option (FMRIB58_FA). I > > then go through all of the remaining steps of TBSS and then when I > > visualise with: > > > > fslview MNI152 mean_FA_skeleton -l Green -b 2000,8000 tbss_tstat1 - > > l Red-Yellow -b 3,6 > > tbss_tstat2 -l Blue-Lightblue -b 3,6 > > > > the mean_FA_skeleton significantly overshoots the brain, > > terminating in or even beyond the skull > > (the overall extent of the skeleton is significantly larger that > > the brain). > > > > Any ideas about what might be going on here would be greatly > > appreciated! > > > > Thanks, > > Marc > > > ------------------------------------------------------------------------ > --- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > ------------------------------------------------------------------------ > --- > -- Marc Dubin, MD PhD Resident in Psychiatry - PGY4 Weill Cornell Medical College 212-746-3764 (w) 646-831-8886 (c) ------=_Part_25261_26458503.1192509497562 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi,

The overlay of the individual FA files with FMRIB58_FA seems to be good but there may be a problem with all_FA - there is a large part of the frontal lobe bilaterally that got cut out, I think during the nonlinear transformation to FMRIB58_FA. The 31 files for the 31 individual subjects are included as well as all_FA in 542195.

Best,
Marc

On 10/15/07, Steve Smith <[log in to unmask] > wrote:
Hi - I'm suprised that you get this with the -T option. If you load
in all_FA first into FSLView, and then overlay the FMRIB58_FA
(threshold this, change its colour to red-yellow, and make it a
little transparent) and then move through the timepoints, do all
subjects behave in a similar way to what you've described?

We can have a look and advise on what's happening;
Please upload the files in a single compressed tarfile to
http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi
and then email me the upload ID.

Cheers.



On 15 Oct 2007, at 02:36, Marc Dubin wrote:

> Hello FSL Users,
>
> Using the TBSS tools I have been carrying out the tbss_2_reg step
> with the -T option (FMRIB58_FA). I
> then go through all of the remaining steps of TBSS and then when I
> visualise with:
>
> fslview MNI152 mean_FA_skeleton -l Green -b 2000,8000 tbss_tstat1 -
> l Red-Yellow -b 3,6
> tbss_tstat2 -l Blue-Lightblue -b 3,6
>
> the mean_FA_skeleton significantly overshoots the brain,
> terminating in or even beyond the skull
> (the overall extent of the skeleton is significantly larger that
> the brain).
>
> Any ideas about what might be going on here would be greatly
> appreciated!
>
> Thanks,
> Marc


------------------------------------------------------------------------
---
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director,  Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222726  (fax 222717)
[log in to unmask]    http://www.fmrib.ox.ac.uk/ ~steve
------------------------------------------------------------------------
---



--
Marc Dubin, MD PhD
Resident in Psychiatry - PGY4
Weill Cornell Medical College

212-746-3764 (w) 646-831-8886 (c)
------=_Part_25261_26458503.1192509497562-- ========================================================================= Date: Tue, 16 Oct 2007 00:45:16 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Marc Dubin <[log in to unmask]> Subject: Re: mean_FA_skeleton and brain misalignment Comments: cc: Steve Smith <[log in to unmask]>, [log in to unmask] In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_25292_6300589.1192509916085" ------=_Part_25292_6300589.1192509916085 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Correction...............the upload reference number is 351859 On 10/15/07, Steve Smith <[log in to unmask]> wrote: > > Hi - I'm suprised that you get this with the -T option. If you load > in all_FA first into FSLView, and then overlay the FMRIB58_FA > (threshold this, change its colour to red-yellow, and make it a > little transparent) and then move through the timepoints, do all > subjects behave in a similar way to what you've described? > > We can have a look and advise on what's happening; > Please upload the files in a single compressed tarfile to > http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi > and then email me the upload ID. > > Cheers. > > > > On 15 Oct 2007, at 02:36, Marc Dubin wrote: > > > Hello FSL Users, > > > > Using the TBSS tools I have been carrying out the tbss_2_reg step > > with the -T option (FMRIB58_FA). I > > then go through all of the remaining steps of TBSS and then when I > > visualise with: > > > > fslview MNI152 mean_FA_skeleton -l Green -b 2000,8000 tbss_tstat1 - > > l Red-Yellow -b 3,6 > > tbss_tstat2 -l Blue-Lightblue -b 3,6 > > > > the mean_FA_skeleton significantly overshoots the brain, > > terminating in or even beyond the skull > > (the overall extent of the skeleton is significantly larger that > > the brain). > > > > Any ideas about what might be going on here would be greatly > > appreciated! > > > > Thanks, > > Marc > > > ------------------------------------------------------------------------ > --- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > ------------------------------------------------------------------------ > --- > -- Marc Dubin, MD PhD Resident in Psychiatry - PGY4 Weill Cornell Medical College 212-746-3764 (w) 646-831-8886 (c) ------=_Part_25292_6300589.1192509916085 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Correction...............the upload reference number is 351859

On 10/15/07, Steve Smith <[log in to unmask]> wrote:
Hi - I'm suprised that you get this with the -T option. If you load
in all_FA first into FSLView, and then overlay the FMRIB58_FA
(threshold this, change its colour to red-yellow, and make it a
little transparent) and then move through the timepoints, do all
subjects behave in a similar way to what you've described?

We can have a look and advise on what's happening;
Please upload the files in a single compressed tarfile to
http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi
and then email me the upload ID.

Cheers.



On 15 Oct 2007, at 02:36, Marc Dubin wrote:

> Hello FSL Users,
>
> Using the TBSS tools I have been carrying out the tbss_2_reg step
> with the -T option (FMRIB58_FA). I
> then go through all of the remaining steps of TBSS and then when I
> visualise with:
>
> fslview MNI152 mean_FA_skeleton -l Green -b 2000,8000 tbss_tstat1 -
> l Red-Yellow -b 3,6
> tbss_tstat2 -l Blue-Lightblue -b 3,6
>
> the mean_FA_skeleton significantly overshoots the brain,
> terminating in or even beyond the skull
> (the overall extent of the skeleton is significantly larger that
> the brain).
>
> Any ideas about what might be going on here would be greatly
> appreciated!
>
> Thanks,
> Marc


------------------------------------------------------------ ------------
---
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director,  Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222726  (fax 222717)
[log in to unmask]    http://www.fmrib.ox.ac.uk/~steve
------------------------------------------------------------------------
---



--
Marc Dubin, MD PhD
Resident in Psychiatry - PGY4
Weill Cornell Medical College

212-746-3764 (w) 646-831-8886 (c) ------=_Part_25292_6300589.1192509916085-- ========================================================================= Date: Tue, 16 Oct 2007 06:39:48 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: fslvm32 In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - there were bugs in 64-bit-vmware that have prevented this so far, but at some point presumably this will get fixed. Cheers. On 14 Oct 2007, at 16:18, Emily RUBIN-FERREIRA wrote: > Great thanks very much for the response. I'm sure we'll switch to > Linux > but in a last ditch effort to continue using windows, are there are > any > plans to come out with fslvm64? > Thanks, > emily > > > On Sun, 14 Oct 2007, Steve Smith wrote: > >> I think that;s right, yes - Duncan can correct me if I'm wrong - I >> doubt this will give you 64-bit functionality. If you are needing to >> do large analyses we would strongly recommend against using Windows >> anyway - I would recommend properly native Linux or Apple. >> >> Cheers, Steve. >> >> >> On 13 Oct 2007, at 17:43, Emily RUBIN-FERREIRA wrote: >> >>> Hi, >>> I would like to install FSL 4.0 on a Windows XP computer 64-bit, >>> using >>> vmware and fslvm32 as instructed. >>> >>> I was reading in the FAQ the following: that on 32-bit computers, >>> there >>> is a limit on the size that a single program can be regardless of >>> the >>> RAM/swap you have, and that the limit is 2 GB. If fslvm is 32bit, >>> does >>> this mean that if I install using fslvm, that my computer becomes >>> like a >>> 32-bit computer? >>> >>> Thanks, >>> emily >> >> >> --------------------------------------------------------------------- >> --- >> --- >> Stephen M. Smith, Professor of Biomedical Engineering >> Associate Director, Oxford University FMRIB Centre >> >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK >> +44 (0) 1865 222726 (fax 222717) >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve >> --------------------------------------------------------------------- >> --- >> --- >> ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Tue, 16 Oct 2007 06:41:04 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: group melodic and individual IC effect sizes In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - yes, that's right. cheers. On 15 Oct 2007, at 22:25, Michelle Voss wrote: > hello, > > i've run a group melodic analysis, and i want to correlate the > extent to which subjects exhibit a component (say the RSN) with > another variable (say age). this info is captured in the subjects/ > session mode plot for individuals about the group mean, and in the > box plot to show formally the group variation about the mean--in > that-- the extent to which a component is represented and regions > in it functionally covary for an individual is represented by their > individual effect size and for a group by mean effect size. when > you contrast two groups (say male and female) within the group PICA > analysis, this contrasts the difference in mean effect (or > representation of that group) for a component. > > is my understanding of the melodic report right? i.e., the > individual effects sizes in the subjects/session mode are > meaningful in looking for individual differences in exemplifying a > component (?) > > thanks so much, > > Michelle > > > > ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Tue, 16 Oct 2007 06:43:04 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: WARNING::in calculating COG, total = 0.0 In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, It's possible that there's something weird in the intensity values or the header. What is the output of running: fslhd fslstats -r -R -v -V -c -C Cheers. On 15 Oct 2007, at 23:01, Kimberly Carpenter wrote: > Hello, > I am having problems running 1st level analyses on a few of my > subjects. > During the prestats stage of the analysis I get a repeated error of > > WARNING::in calculating COG, total = 0.0 > > I see that previous posts have said that "The COG total=0 warning > occurs > when the images are empty - that is, all zeros." However, when I > look at my > raw data using a Matlab script I have images. Does anyone know > what may be > causing this error? Thanks! > ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Tue, 16 Oct 2007 06:44:51 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: RE : [FSL] TBSS - file exclusion In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=ISO-8859-1; delsp=yes; format=flowed Content-Transfer-Encoding: quoted-printable Indeed - or you could just remove all files relating to the bad =20 subject from FAi and then rerun tbss_3_postreg etc. Cheers. On 15 Oct 2007, at 19:49, Gwena=EBlle DOUAUD wrote: > Hi Conrad, > > I'm not sure if somebody has already answered your > question, it seems that I receive all the emails quite > delayed today. Anyway, sure, if you can not fix it, > you can always remove this image from your 4D dataset. > If the bad scan is the volume #i in _fslview_ among > your n images, you just need to use fslroi with > something like: > fslroi all_FA all_FA_1 0 i > fslroi all_FA all_FA_2 i+1 n-(i+1) > fslmerge -t all_FA all_FA_1 all_FA_2 > > Also, please just make sure that the badly registered > scan has not mislead the mean_FA and mean_FA_skeleton > creation (hopefully not). > > Hope this helps, > Gwenaelle > > > --- Conrad Rockel <[log in to unmask]> a =E9crit : > >> Hi there, >> >> I am working through the various TBSS stages, and >> upon reviewing the output >> of tbss_3_postreg, I noticed that one of the scans >> is badly registered. Is >> there a way I can remove this scan from the 4D image >> and subsequent analysis >> without having to run the previous step again >> (because it took several days >> to run)? >> >> Cheers, >> >> Conrad >> > > > > =20 > ______________________________________________________________________=20= > _______ > Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers =20 > Yahoo! Mail ------------------------------------------------------------------------=20= --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------=20= --- ========================================================================= Date: Tue, 16 Oct 2007 06:57:34 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: recreating varcopes In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi - I suspect that either the difference lies in the exact equation that you're using (see for example eq 14.9 in Keith's chapter in our FMRI book), or some implementational detail, in which case Wooly will need to help here. Cheers, Steve. On 16 Oct 2007, at 05:33, Jeanette Mumford wrote: > Hi, > > I was just trying to recreate the varcope image from a first level > analysis using the sigmasquareds and threshac1 images as well as the > design.mat file. My result is close, but I'm wondering why it isn't > closer. Here's what I've done > > 1. Create the correlation matrix V using the first element of > threshac1 as the main diagonal, 2nd element as the first diagonal > above and below the main diagonal and so forth. > 2. Demean the design matrix that's stored in design.mat (call it X) > 3. Calculate inv(X'*inv(V)*X)*sigma^2 Where sigma^2 is the > sigmasquareds value for that voxel and take the appropriate contrast > of this matrix to correspond with my varcope of interest. > > My method seems to produce a variance that is consistently smaller > than the varcope, but within 20% or so. Did I miss something? > Highpass filtering was used in the analysis (both the data and the > design), but I figured that would already be incorporated in all the > pieces that I used. The analysis was originally carried out in Feat > version 5.43. > > Thanks! > Jeanette ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Tue, 16 Oct 2007 07:00:14 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: MNI Space in MRIcro and FSL In-Reply-To: [log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=WINDOWS-1252; delsp=yes; format=flowed Content-Transfer-Encoding: quoted-printable HI - I presume this is because of a change in field-of-view in the =20 different standard space images. One easy way to solve that is to use =20= flirt to estimate the transform between the standard space image in =20 SPM/mricro and the FSL standard space image, and use that instead of =20 eye.mat in your call to flirt resample the ROI. Cheers, Steve. On 16 Oct 2007, at 00:17, Matt Glasser wrote: > Hi, > > > > I have a set of ROIs from SPM in MNI space that I want to load into =20= > FSL. When I take the standard FSL brain MNI152_T1_1mm_brain and =20 > load it in MRIcro and overlay the ROI, it looks fine, like the =20 > first screenshot: http://www.ma-tea.com/~Matt/MRIcroOnMNI.png =20 > However, if I resample the 4X4X4mm ROI so it can be overlayed on =20 > this brain in FSLView, using: > > > > flirt =96in 4X4X4mmROI =96ref MNI152_T1_1mm_brain =96applyxfm =96init =20= > eye.mat =96out 1X1X1mmROI where eye.mat is the identity matrix: > > > > 1 0 0 0 > > 0 1 0 0 > > 0 0 1 0 > > 0 0 0 1 > > > > The result is incorrect, shown by the second screenshot in FSLView: =20= > http://www.ma-tea.com/~Matt/FSLOnMNI.png > > > > The ROI file header is the following: > > > > > nifti_type =3D 'ANALYZE-7.5' > > image_offset =3D '0' > > ndim =3D '4' > > nx =3D '41' > > ny =3D '48' > > nz =3D '35' > > nt =3D '1' > > dx =3D '4' > > dy =3D '4' > > dz =3D '4' > > dt =3D '1' > > datatype =3D '16' > > nbyper =3D '4' > > byteorder =3D 'LSB_FIRST' > > cal_min =3D '0' > > cal_max =3D '1' > > descrip =3D 'FSL4.0' > > sform_code =3D '2' > > sto_xyz_matrix =3D '-4 0 0 80 0 4 0 -112 0 0 4 -52 0 0 0 1' > > num_ext =3D '0' > > /> > > > > I would appreciate any assistance with this as I am unfamiliar with =20= > this type of issue if it has to do with the analyze file format or =20 > header. > > > > Thanks, > > > > Matt. > > ------------------------------------------------------------------------=20= --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------=20= --- ========================================================================= Date: Tue, 16 Oct 2007 07:01:52 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: RE : [FSL] The orientation problem in VBM analysis In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=UTF-8; delsp=yes; format=flowed Content-Transfer-Encoding: quoted-printable Hi - you need to try different orderings of x,y,z (including possibly =20= putting '-' in front of each of these) inside the fslswapdim call. =20 Whenever I do this I just keep trying things until it =20 works.....faster than trying to think it through.... Cheers. On 15 Oct 2007, at 23:50, Zhang, Xiaochu (NIH/NIDA) [F] wrote: > Thank you very much for your help! I read the mail you gave to me. =20 > However, I found we have different problems. In our case, we don=E2=80=99= t =20 > want to only change L/R. However, We want to change the y axial and =20= > the z axial. Now, the z (y) axial should be A/P (S/I) but is S/I (A/=20= > P) now. I changed the orientation in NIFTI via "fslorient". =20 > However, after "tbss_1_preprocess", it is wrong, again. The =20 > "fslorient" looks seemly not work. > The "fslswapdim" is worked. However, I don't know how to change y =20 > axial and z axial by it. > Could you help me? I'm eager to your help. > Thanks, again! > Xiaochu > > -----Original Message----- > From: Gwena=C3=ABlle DOUAUD [mailto:[log in to unmask]] > Sent: 2007=E5=B9=B410=E6=9C=8813=E6=97=A5 22:05 > To: [log in to unmask] > Subject: [FSL] RE : [FSL] The orientation problem in VBM analysis > > Hi again Xiaochu, > > for this orientational problem, please look at > http://www.jiscmail.ac.uk/cgi-bin/webadmin?=20 > A2=3Dind0709&L=3DFSL&P=3DR44248&I=3D-3 > on the same subject... > > Cheers, > Gwena=C3=ABlle > > --- "Zhang, Xiaochu (NIH/NIDA) [F]" > <[log in to unmask]> a =C3=A9crit : > >> Hi experts, >> >> I am working on the VBM analysis. For the data is >> from multiple studies, >> some is sagitial and some is axial. After "tbss_1" >> and check the >> slicedir, the data showed inconsistent orientation. >> Who can help me and >> tell me how to change the orientation? >> >> Thanks a lot! >> >> >> >> Xiaochu Zhang Ph.D >> >> Visiting Research Fellow >> >> NIH/NIDA-IRP >> >> 5500 Nathan Shock Drive >> >> Baltimore MD 21224 >> >> >> >> Tel: (410) - 550 - 1440 ext. 434 >> >> >> >> > > > > =20 > ______________________________________________________________________=20= > _______ > Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers =20 > Yahoo! Mail ------------------------------------------------------------------------=20= --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------=20= --- ========================================================================= Date: Tue, 16 Oct 2007 07:04:30 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: Using SIENA in monkey MRIs In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, I would NOT realign the images before running SIENA/FAST etc. This is because doing this reduces the resolution of the image because of interpolation-related blurring. It's ok to use fslswapdim as that just switches the axes around without having to resample between voxels. Otherwise, yes, just send the whole SIENA output directory. Cheers. On 15 Oct 2007, at 17:48, Buyean Lee wrote: > Hi Steve, > > Would you let me know which files you want to see; there are three > kinds of MRIs? > > 1. Analyze file formats in the orientation of acquisition (because > the nose is too big, I couldn't use AC-PC alignment). > 2. Analyze file formats after AC-PC alignment; resliced images > 3. Nifti file formats after removing the neck and tissues outside > the skull (these are the files works best in SIENA). > > I think it might be informative if I can send you the whole > directory SIENA created so that you can see the report. > Please let me know if you want me to do it. > > Thank you, > > Buyean > > -----Original Message----- > From: Steve Smith <[log in to unmask]> > To: [log in to unmask] > Sent: Sun, 14 Oct 2007 3:35 am > Subject: Re: [FSL] Using SIENA in monkey MRIs > > Hi, > > Is the movement between the two timepoints after registration due > to atrophy effects of interest, or "real" misregistration? > > If you would like us to have a quick look, please upload the files > in a single compressed tarfile to > http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi > and then email me the upload ID. > > Cheers. > > > On 11 Oct 2007, at 03:15, Buyean Lee wrote: > > > Hi Steve, > > > > Finally, I ran Siena with the monkey MRIs (it is just one > subject); > the siena_report.html looks much better in terms of > brain > extraction, coregistation, etc. > > > > This is what I did. > > > > 1. removed all tissues outside the skull and below the cerebellum. > > 2. modified the siena.sh in order to use separate BET option for > > two scans. > > > > Brain extraction is much better, but coregistration by FLIRT is > not > satisfactory. > > I can still see some movement in the overlap of the two > coregistred > MRIs. > > > > Is there any way for me to send these two MRIs to you so that you > > can take a look at them? > > > > I just don't know what else I can do to improve the BET and > > coregistration further. > > > > Thank you, > > > > Buyean > > > > > > > > -----Original Message----- > > From: Steve Smith <[log in to unmask]> > > To: [log in to unmask] > > Sent: Thu, 27 Sep 2007 1:20 am > > Subject: Re: [FSL] Using SIENA in monkey MRIs > > > > HI - you would also want to include a rasonable amount of the > outer > skull boundary that's near the brain boundary, as this gets > used in > SIENA. > > > > To choose a FOV in FSLView and then reduce the image by this: > > Click around and find the voxel coordinates (voxel not mm) that > > bound the region you want to keep. > > Find the xmin and xmax values, and calculate the xsize = xmax - xmin > > Same for y and z. > > > > Then run > > > > fslroi originalheadimage reducedimage xmin xsize ymin ysize zmin > zsize > > > > (do this for both timepoints separately) > > > > and feed the reducedimages into SIENA. > > > > > > Cheers. > > > > On 26 Sep 2007, at 23:52, Buyean Lee wrote: > > > > > Hi Steve, > > > > > > "Alternatively, you could work out how to reduce the field-of-> > view > to just include the brain of each time point before feeding > > the > output of that into SIENA. You would use FSLView on each > > image to > work out the field-of-view and then use fslroi to reduce > > the FOV of > the image." > > > > > > I created a whole brain mask for the animal brain using FSLView. > > > > > > But, it is not easy to understand how to use fslroi. > > > > > > fslroi ... > > > > > > I just don't know how to determine these parameters and when I > am > > supposed to use the mask file in fslroi. > > > > > > Would you kindly let me know how to reduce the field-of-view to > > > include the brain? > > > > > > > > > Thank you, > > > > > > Buyean > > > > > ---------------------------------------------------------------------- > > > ----- > > > Check Out the new free AIM(R) Mail -- Unlimited storage and > > > industry-leading spam and email virus protection. > > > > > ---------------------------------------------------------------------- > > ----- > > Stephen M. Smith, Professor of Biomedical Engineering > > Associate Director, Oxford University FMRIB Centre > > > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > > +44 (0) 1865 222726 (fax 222717) > > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > > > ---------------------------------------------------------------------- > > ----- > > Check Out the new free AIM(R) Mail -- Unlimited storage and > > industry-leading spam and email virus protection. > > ---------------------------------------------------------------------- > ----- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > ---------------------------------------------------------------------- > ----- > Check Out the new free AIM(R) Mail -- Unlimited storage and > industry-leading spam and email virus protection. ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Tue, 16 Oct 2007 07:18:01 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: DTI: noise estimation In-Reply-To: <[log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, On 15 Oct 2007, at 16:22, Najmeh Khalili M. wrote: > Hi, > > Is there a way to get the residuals of eddy_correction or is > there any way to estimate a metric of suseptibility distortions > present in DTI? Eddy_correction is not a model-fitting (in the way that dtifit is), so you don't get "residuals" - however, you can look at the temporal variance of the uncorrected and corrected datasets, though remember that this includes variation due to the real valid diffusion effects. To do it on either 4D dataset: fslmaths <4Ddata> -Tstd dataSTD However the majority of the variance you see here will not be related to susceptibility distortions. The geometric distortions would be modelled, for example, by using the FUGUE tool and a fieldmap (see the manual for more details). > More generally, I need to model the imaging noise in the > between-subject DTI analysis and I wonder if you have any > tools < or philosophy > that address this problem. Again - I'm not sure this I follow - this is another separate issue - the majority of between-subject variation will probably be neither due to susceptibility distortions nor within-subject acquisition noise, but due to real subject-subject variability in tract geometry and FA values. Cheers. > > Thank you, > Naj ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Tue, 16 Oct 2007 07:58:58 -0400 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Matt Glasser <[log in to unmask]> Organization: ma-tea.com Subject: Re: MNI Space in MRIcro and FSL In-Reply-To: <[log in to unmask]> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Hi Steve, Well okay, but why would the ROIs line up with FSL's standard brain in MRIcro, but not in FSL itself? I used the same standard space image in both programs and it was only misaligned in one of them, so I don't know how I could use a flirt transform in this case. Thanks, Matt. -----Original Message----- From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Steve Smith Sent: Tuesday, October 16, 2007 2:00 AM To: [log in to unmask] Subject: Re: [FSL] MNI Space in MRIcro and FSL HI - I presume this is because of a change in field-of-view in the different standard space images. One easy way to solve that is to use flirt to estimate the transform between the standard space image in SPM/mricro and the FSL standard space image, and use that instead of eye.mat in your call to flirt resample the ROI. Cheers, Steve. On 16 Oct 2007, at 00:17, Matt Glasser wrote: > Hi, > > > > I have a set of ROIs from SPM in MNI space that I want to load into > FSL. When I take the standard FSL brain MNI152_T1_1mm_brain and > load it in MRIcro and overlay the ROI, it looks fine, like the > first screenshot: http://www.ma-tea.com/~Matt/MRIcroOnMNI.png > However, if I resample the 4X4X4mm ROI so it can be overlayed on > this brain in FSLView, using: > > > > flirt -in 4X4X4mmROI -ref MNI152_T1_1mm_brain -applyxfm -init > eye.mat -out 1X1X1mmROI where eye.mat is the identity matrix: > > > > 1 0 0 0 > > 0 1 0 0 > > 0 0 1 0 > > 0 0 0 1 > > > > The result is incorrect, shown by the second screenshot in FSLView: > http://www.ma-tea.com/~Matt/FSLOnMNI.png > > > > The ROI file header is the following: > > > > > nifti_type = 'ANALYZE-7.5' > > image_offset = '0' > > ndim = '4' > > nx = '41' > > ny = '48' > > nz = '35' > > nt = '1' > > dx = '4' > > dy = '4' > > dz = '4' > > dt = '1' > > datatype = '16' > > nbyper = '4' > > byteorder = 'LSB_FIRST' > > cal_min = '0' > > cal_max = '1' > > descrip = 'FSL4.0' > > sform_code = '2' > > sto_xyz_matrix = '-4 0 0 80 0 4 0 -112 0 0 4 -52 0 0 0 1' > > num_ext = '0' > > /> > > > > I would appreciate any assistance with this as I am unfamiliar with > this type of issue if it has to do with the analyze file format or > header. > > > > Thanks, > > > > Matt. > > ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ========================================================================= Date: Tue, 16 Oct 2007 13:27:28 +0100 Reply-To: FSL - FMRIB's Software Library <[log in to unmask]> Sender: FSL - FMRIB's Software Library <[log in to unmask]> From: Steve Smith <[log in to unmask]> Subject: Re: MNI Space in MRIcro and FSL In-Reply-To: [log in to unmask]> Mime-Version: 1.0 (Apple Message framework v752.3) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Content-Transfer-Encoding: 7bit Hi, On 16 Oct 2007, at 12:58, Matt Glasser wrote: > Hi Steve, > > Well okay, but why would the ROIs line up with FSL's standard brain in > MRIcro, but not in FSL itself? I used the same standard space > image in both > programs Ah - I hadn't appreciated that - sorry... > and it was only misaligned in one of them, so I don't know how I > could use a flirt transform in this case. Right - in that case the difference presumably is because MRICRO uses the co-ordinate centre from the header in determining display centering, whereas FSLView only uses the raw voxels for working out display (the co-ordinate centre is only used in the reporting of the mm co-ordinates, not in the image display). Cheers, Steve. > > Thanks, > > Matt. > > -----Original Message----- > From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On > Behalf > Of Steve Smith > Sent: Tuesday, October 16, 2007 2:00 AM > To: [log in to unmask] > Subject: Re: [FSL] MNI Space in MRIcro and FSL > > HI - I presume this is because of a change in field-of-view in the > different standard space images. One easy way to solve that is to use > flirt to estimate the transform between the standard space image in > SPM/mricro and the FSL standard space image, and use that instead of > eye.mat in your call to flirt resample the ROI. > > Cheers, Steve. > > > On 16 Oct 2007, at 00:17, Matt Glasser wrote: > >> Hi, >> >> >> >> I have a set of ROIs from SPM in MNI space that I want to load into >> FSL. When I take the standard FSL brain MNI152_T1_1mm_brain and >> load it in MRIcro and overlay the ROI, it looks fine, like the >> first screenshot: http://www.ma-tea.com/~Matt/MRIcroOnMNI.png >> However, if I resample the 4X4X4mm ROI so it can be overlayed on >> this brain in FSLView, using: >> >> >> >> flirt -in 4X4X4mmROI -ref MNI152_T1_1mm_brain -applyxfm -init >> eye.mat -out 1X1X1mmROI where eye.mat is the identity matrix: >> >> >> >> 1 0 0 0 >> >> 0 1 0 0 >> >> 0 0 1 0 >> >> 0 0 0 1 >> >> >> >> The result is incorrect, shown by the second screenshot in FSLView: >> http://www.ma-tea.com/~Matt/FSLOnMNI.png >> >> >> >> The ROI file header is the following: >> >> >> >> > >> nifti_type = 'ANALYZE-7.5' >> >> image_offset = '0' >> >> ndim = '4' >> >> nx = '41' >> >> ny = '48' >> >> nz = '35' >> >> nt = '1' >> >> dx = '4' >> >> dy = '4' >> >> dz = '4' >> >> dt = '1' >> >> datatype = '16' >> >> nbyper = '4' >> >> byteorder = 'LSB_FIRST' >> >> cal_min = '0' >> >> cal_max = '1' >> >> descrip = 'FSL4.0' >> >> sform_code = '2' >> >> sto_xyz_matrix = '-4 0 0 80 0 4 0 -112 0 0 4 -52 0 0 0 1' >> >> num_ext = '0' >> >> /> >> >> >> >> I would appreciate any assistance with this as I am unfamiliar with >> this type of issue if it has to do with the analyze file format or >> header. >> >> >> >> Thanks, >> >> >> >> Matt. >> >> > > > ---------------------------------------------------------------------- > -- > --- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > ---------------------------------------------------------------------- > -- > --- ------------------------------------------------------------------------ --- Stephen M. Smith, Professor of Biomedical Engineering Associate Director, Oxford University FMRIB Centre FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK +44 (0) 1865 222726 (fax 222717) [log in to unmask] http://www.fmrib.ox.ac.uk/~steve ------------------------------------------------------------------------ --- ======================================