Do you mean the 4d "grand mean" intensity scaling? You can't turn this
off from the GUI (it's always useful to do and has no within-session
effect on the results).
The relevant line in the log is
/usr/local/fsl/bin/fslmaths prefiltered_func_data_smooth -mul
1.46139553856 prefiltered_func_data_intnorm
Where the number will be data-dependent.
cheers.
On 15 Jan 2009, at 22:44, Jee Eun Lee wrote:
> Hello,
>
> I'd like to turn off this global intensity scaling feature. How can
> I do that in the Feat GUI? or which command line does do it in the
> log file?
>
> Thank you,
>
> Jee Eun
>
>
>
> On Sun, Jan 4, 2009 at 3:12 AM, Mark Jenkinson <[log in to unmask]>
> wrote:
> Hi,
>
> If your intensities are high everywhere then it is just the global
> intensity
> scaling which FEAT does on the data. This is a single constant
> applied
> to all voxels and so will not affect any of the DTI measures.
>
> The processing of your fieldmap sounds fine, so everything sounds OK
> to now use your unwarped images for further DTI processing.
>
> All the best,
> Mark
>
>
>
> On 2 Jan 2009, at 18:36, Rajagopalan, Venkateswaran wrote:
>
> Dear Mark,
>
> Thanks for your kind response. Intensities were high (if original
> intensity =145 the final undistorted image has intensity of 19500)
> all over the brain in undistorted final image when compared with the
> original distorted image.I checked in some of the white matter areas
> as per your guidelines they are very high. I will give you the
> details about my field map image and how i went step by step.
>
> Field maps were acquired with Gradient echo sequence i have two
> magnitude images and a single field map image which the physicist
> told me is the phase difference image. The intensity range of this
> phase difference image is from -4096 to +4096 so in my first step i
> scaled this range to -pi to +pi then i did phase unwrapping using
> PRELUDE and then i converted this image to rad/sec dividing by
> difference in TEs. Then i submitted this image to FUGUE GUI and fed
> in appropriate parameters.
>
> Thanks again for you information on FWHM and other questions.
>
> Regards
> venkateswaran
>
>
> -----Original Message-----
> From: FSL - FMRIB's Software Library on behalf of Mark Jenkinson
> Sent: Fri 1/2/2009 9:13 AM
> To: [log in to unmask]
> Subject: Re: [FSL] Fugue -field map correction
>
> Hi,
>
> I'm concerned about the change in intensity range. A small change I
> would expect
> but nothing like this magnitude. Do you see this kind of change in
> many voxels or is there
> just one outlier peak voxel somewhere pushing up the range? Have a
> look at the
> typical intensities in an area of white matter (with FSLView) before
> and after the correction.
> If there is a big change in intensity here then we need to figure out
> why. If it doesn't seem
> to change then maybe it is only an outlier voxel. In that case, look
> at the 99th percentile
> (with fslstats) and see how that changes. If you do have any large
> changes it will certainly
> affect your FA, etc.
>
> As for FWHM, we have implemented an alternative in the FEAT GUI which
> only regularises
> at the boundary. We have found that this gives the desired benefits
> (suppressing noise in
> the voxels that are most affected by it) without having to smooth
> areas where the SNR is good.
> So there is no FWHM equivalent when doing it via the FEAT GUI.
> However, if you find that
> your fieldmaps benefit from smoothing then you can do some smoothing
> (via the command
> line fugue, or with fslmaths) before using it with the FEAT GUI.
>
> All the best,
> Mark
>
>
> On 29 Dec 2008, at 15:38, Rajagopalan, Venkateswaran wrote:
>
> Dear Mark,
>
> Thanks for your kind help. I followed your guidelines and got good
> results i.e my dti images are undistorted. Your advise to use report
> page and FSL's course material on FUGUE was very helpful to find out
> my mistakes. I made the following mistake that is i didn't use
> PRELUDE to unwrap the phase in the preliminary processing stage as
> mentioned in the FSL's FUGUE course material that solved the
> problem, i thought the PRELUDE step is also included in the FUGUE
> GUI option. There is only one problem which i would like to clarify
> with you, before that, i would like to mention that i followed
> exactly the same steps as mentioned in FSL's FUGUE course website
> where i first converted my phase image to -pi to +pi range then
> applied PRELUDE to unwrap it, then converted to rad/sec and then fed
> the field map to pre stats part of FUGUE GUI and set the settings
> there as per your guidelines. My final result image
> filtered_func_data has intensities which range from 0 to 271021
> whereas my original DTI image before field map correction has
> intensities from 0 to 2825 i understand that intensities will be
> modified in my fieldmap corrected image due to so many processing
> steps but my question is; is this high intensity value changes
> common after field map correction or am i making any mistake and
> will this high intensity values affect DTI parameters like FA, RA
> and other DTI metrics which i would like to calculate after this
> distortion correction step.
> One other question is, in FUGUE GUI i didn't choose a value for FWHM
> explicitly as i fed values for dwell time and % signal threshold
> values, how is this FWHM value chosen whether the algorithm
> calculates the appropriate value by using some kind of iterative
> procedure.
> If i have not given you enough information kindly inform me, i will
> get back to you with details.
>
> Thanks a lot for your kind help. Wishing you a Very Happy New Year. I
>
> Thanks
>
> Regards
> venkateswaran
>
>
>
> -----Original Message-----
> From: FSL - FMRIB's Software Library on behalf of Mark Jenkinson
> Sent: Tue 12/23/2008 5:48 PM
> To: [log in to unmask]
> Subject: Re: [FSL] Fugue -field map correction
>
> Hi,
>
> You are right about the units of dwell time - in the GUI they are in
> ms, so 0.85 should be OK.
> What you are doing sounds generally right, although I would turn off
> spatial smoothing as well.
> You don't need to specify a TR or a highpass filter cutoff for this
> application (you do need to
> turn the filtering off on the pre-stats tab).
>
> For signal loss you could go down to 0% for spin-echo EPI, but it
> won't really hurt
> to have it at 10%. You shouldn't need to do registration either. So
> basically it all sounds
> fine except for a bit of extra/unnecessary smoothing. I would expect
> you to get reasonable
> results and don't understand why you get excessively large
> intensities. What other things
> go wrong? Does it apply too strong a change, or not enough, or the
> wrong direction, or what?
>
> Are you also looking at the FEAT output report (the report.html page
> and the pre-stats output)?
> This will help diagnose whether there is a problem in registration or
> field-map preparation or
> direction of warping. Look at the fsl course practical on FUGUE
> (under registration) for more help.
> http://www.fmrib.ox.ac.uk/fslcourse/lectures/practicals/reg/index.html#fugue
>
> If you still cannot figure out what is going wrong then let us know.
> All the best,
> Mark
>
>
>
> On 23 Dec 2008, at 21:06, Rajagopalan, Venkateswaran wrote:
>
> Dear Matt,
>
> Thanks, i followed exactly the commands you described when i used
> command line approach that is yours i divided my dwell time = 0.85
> msec by 1000 as you showed, whereas in GUI mode i have to enter them
> in msec only. Thanks for your kind infromation.
>
> Regards
> venkateswaran
>
> ________________________________
>
> From: FSL - FMRIB's Software Library on behalf of Matt Glasser
> Sent: Tue 12/23/2008 3:47 PM
> To: [log in to unmask]
> Subject: Re: [FSL] Fugue -field map correction
>
>
>
> Your dwell time seems rather high compared to the one I used.
> Perhaps you
> need to divide it by 1000 and try again? I think it is measured in
> milliseconds but fugue takes it in seconds.
>
> Peace,
>
> Matt.
>
> -----Original Message-----
> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On
> Behalf
> Of Rajagopalan, Venkateswaran
> Sent: Tuesday, December 23, 2008 1:31 PM
> To: [log in to unmask]
> Subject: Re: [FSL] Fugue -field map correction
>
> Dear Mark,
>
> Thanks for your kind information. I tried using Pre stats in FEAT
> GUI as per
> your guidelines but I didn't get good results. I would like clarify
> with you
> some of my doubts; 1) i chose only "pre-stats " from the top pull
> down menu
> as a result of this the icons in the bottom which were enabled to do
> pre
> stats operation are Misc, Data, Pres-stats and Registration, i
> unchecked the
> option Balloon help and progress watcher under the "Misc" option.
> Then under
> the data icon i loaded my DTI data set and chose an output folder to
> save my
> results, here under the "Data" icon do i need to give the value of
> TR for my
> EPI sequence and do i also need to specify the high pass filter
> cutoff
> value. Then under "Pres stats" icon i chose B0 unwarping option and
> unchecked all other options except FWHM like BET extraction ,
> temporal
> filtering etc then gave my field map image and BET extracted
> magnitude
> image, dwell time and EPI TE time and chose my unwarp direction, i
> don't
> know how to choose the %signal loss threshold any criteria available
> for
> that currently i retained the default value of 10, then
> corresponding to
> sigma of 0.5 i chose FWHM =2.35*0.5 ~1.2. Then i unchecked the
> default
> checked Linear registration icon, is it correct to uncheck this
> registration
> option or i have to leave it checked as such.
> I tried Matt's command line approach also but didn't get good
> results. I
> repeated the above steps by changing the unwrap direction from y to -
> y but
> once again i didn't get good results. One more doubt i want to
> clarify is
> the end result "filtered_func_data" has intensity values which are
> very
> large when compared to the original DTI image intensity whereas i
> didn't see
> these large values in my final image when i followed Matt's approach.
>
> I am sorry to bother you,
>
> Thanks
>
> Regards
> venkateswaran
>
>
>
> -----Original Message-----
> From: FSL - FMRIB's Software Library on behalf of Mark Jenkinson
> Sent: Mon 12/22/2008 5:28 PM
> To: [log in to unmask]
> Subject: Re: [FSL] Fugue -field map correction
>
> Hi,
>
> The pre-processing of the fieldmap sounds fine.
>
> The FEAT GUI is the best way to get fieldmap unwarping done as
> it will do all the necessary registration steps (and some extra
> cleanup
> of the fieldmap) for you automatically. All you need to do is give
> it
> your fieldmap in rad/s, the corresponding magnitude image (after
> running BET on it to remove the non-brain structures) and some
> parameter values (e.g. echo spacing). It is all contained under
> the "pre-stats" part and you can run it on your DTI images by
> just selecting pre-stats only as the main option (top left pull-down
> menu), using your DTI as the 4D input data, and turning off all the
> other pre-stats options. The output file will be called
> filtered_func_data
> but will represent your distortion corrected EPI data.
>
> That should work for you with your data.
> If you have any trouble then let us know.
>
> All the best,
> Mark
>
>
>
> On 22 Dec 2008, at 22:19, Rajagopalan, Venkateswaran wrote:
>
>
> Dear All,
>
> I am new to field map correction. My field map data acquisition is a
> single real field map image, gradient echo sequence. I have the
> following images 1) EPI (for DTI),2)single real field map image in
> units rad/sec and 3) the corresponding gradient echo magnitude
> images. I went through FUGUE instruction manual and also through
> PRELUDE documentation. There are few things which i am not able to
> understand. I went through the field map processing steps mentioned
> here http://www.fmrib.ox.ac.uk/fsl/fugue/feat_fieldmap.html and made
> sure that my field map images meet the requirement for instance my
> field map is a single real field map image registered with the
> magnitude image (step 1b and c ), its units are in rad/sec
> (confirmed with MR technician), i regularised them using gaussian
> FWHM of 0.5 fugue --loadfmap=fieldmap_rads -s 1 --
> savefmap=fieldmap_regu. After this step i used the following FUGUE
> command fugue -i epi --dwell=0.85 --loadfmap=fieldmap_regu -u
> result. My first question is this approach correct.
>
> My Next question is, i have uploaded the screenshot of my "result"
> image in the following link
> http://www.beetony.com/uploads/venk/FUGUE_result.jpg
> where you can see the problem, i feel that this is due to the fact
> that field map and EPI images are not registered. I don't know
> whether FUGUE takes cares of the registration between the gradient
> echo field map which is of dim 63*63*30 and epi (DTI) sequence dim
> 127*127*30 or i have to register my fieldmap by distorting it using
> FUGUE option "fugue -i undistortedimage -p unwrappedphase --
> dwell=dwelltime --asym=asymtime --nokspace -s 0.5 -w warpedimage"
> and then register the two images together and then use the following
> FUGUE option "fugue -i epi --dwell=0.85 --loadfmap=fieldmap_regu -u
> result" where now epi and fieldmap_regu are brought into common
> space.
>
> I am sorry this may be a stupid question or approach kindly help me
> with this.
>
> Thanks in advance,
>
> venkateswaran
>
>
> P Please consider the environment before printing this e-mail
>
>
>
> Cleveland Clinic is ranked one of the top hospitals
> in America by U.S. News & World Report (2008).
> Visit us online at http://www.clevelandclinic.org <http://www.clevelandclinic.org/
> for
> a complete listing of our services, staff and
> locations.
>
>
> Confidentiality Note: This message is intended for use
> only by the individual or entity to which it is addressed
> and may contain information that is privileged,
> confidential, and exempt from disclosure under applicable
> law. If the reader of this message is not the intended
> recipient or the employee or agent responsible for
> delivering the message to the intended recipient, you are
> hereby notified that any dissemination, distribution or
> copying of this communication is strictly prohibited. If
> you have received this communication in error, please
> contact the sender immediately and destroy the material in
> its entirety, whether electronic or hard copy. Thank you.
>
>
>
>
> ===================================
>
> P Please consider the environment before printing this e-mail
>
> Cleveland Clinic is ranked one of the top hospitals
> in America by U.S. News & World Report (2008).
> Visit us online at http://www.clevelandclinic.org <http://www.clevelandclinic.org/
> for
> a complete listing of our services, staff and
> locations.
>
>
> Confidentiality Note: This message is intended for use
> only by the individual or entity to which it is addressed
> and may contain information that is privileged,
> confidential, and exempt from disclosure under applicable
> law. If the reader of this message is not the intended
> recipient or the employee or agent responsible for
> delivering the message to the intended recipient, you are
> hereby notified that any dissemination, distribution or
> copying of this communication is strictly prohibited. If
> you have received this communication in error, please
> contact the sender immediately and destroy the material in
> its entirety, whether electronic or hard copy. Thank you.
>
>
>
>
>
> P Please consider the environment before printing this e-mail
>
> Cleveland Clinic is ranked one of the top hospitals
> in America by U.S. News & World Report (2008).
> Visit us online at http://www.clevelandclinic.org for
> a complete listing of our services, staff and
> locations.
>
>
> Confidentiality Note: This message is intended for use
> only by the individual or entity to which it is addressed
> and may contain information that is privileged,
> confidential, and exempt from disclosure under applicable
> law. If the reader of this message is not the intended
> recipient or the employee or agent responsible for
> delivering the message to the intended recipient, you are
> hereby notified that any dissemination, distribution or
> copying of this communication is strictly prohibited. If
> you have received this communication in error, please
> contact the sender immediately and destroy the material in
> its entirety, whether electronic or hard copy. Thank you.
>
>
>
>
> ===================================
>
> P Please consider the environment before printing this e-mail
>
> Cleveland Clinic is ranked one of the top hospitals
> in America by U.S. News & World Report (2008).
> Visit us online at http://www.clevelandclinic.org for
> a complete listing of our services, staff and
> locations.
>
>
> Confidentiality Note: This message is intended for use
> only by the individual or entity to which it is addressed
> and may contain information that is privileged,
> confidential, and exempt from disclosure under applicable
> law. If the reader of this message is not the intended
> recipient or the employee or agent responsible for
> delivering the message to the intended recipient, you are
> hereby notified that any dissemination, distribution or
> copying of this communication is strictly prohibited. If
> you have received this communication in error, please
> contact the sender immediately and destroy the material in
> its entirety, whether electronic or hard copy. Thank you.
>
>
>
>
>
> P Please consider the environment before printing this e-mail
>
> Cleveland Clinic is ranked one of the top hospitals
> in America by U.S. News & World Report (2008).
> Visit us online at http://www.clevelandclinic.org for
> a complete listing of our services, staff and
> locations.
>
>
> Confidentiality Note: This message is intended for use
> only by the individual or entity to which it is addressed
> and may contain information that is privileged,
> confidential, and exempt from disclosure under applicable
> law. If the reader of this message is not the intended
> recipient or the employee or agent responsible for
> delivering the message to the intended recipient, you are
> hereby notified that any dissemination, distribution or
> copying of this communication is strictly prohibited. If
> you have received this communication in error, please
> contact the sender immediately and destroy the material in
> its entirety, whether electronic or hard copy. Thank you.
>
>
---------------------------------------------------------------------------
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
---------------------------------------------------------------------------
|