On 28 Oct 2007, at 01:00, FSL automatic digest system wrote:
> There are 9 messages totalling 872 lines in this issue.
>
> Topics of the day:
>
> 1. fsl4.0 on Intel Mac doesn't work for SMB disk
> 2. siena, sienax combined with betpremask
> 3. can I implement the following in FEAT
> 4. error in dtifit
> 5. Higher level analysis (2)
> 6. Using SIENA in monkey MRIs
> 7. mask.nii.gz problem in 3rd level feat
> 8. Skull and scalp extraction
>
> ----------------------------------------------------------------------
>
> Date: Sat, 27 Oct 2007 10:24:19 +0100
> From: Steve Smith <[log in to unmask]>
> Subject: Re: fsl4.0 on Intel Mac doesn't work for SMB disk
>
> Hi Brad,
>
> I'm afraid that we definitely do not recommend ever using an SMB-
> mounted file system for anything remotely unix-related. It is not a
> full implementation of a decent file system - for example, file
> permissions and ownership are not handled correctly, plus more hidden
> problems like you are seeing.
>
> Sorry not to be more helpful - but you would be much better off with
> a "proper" unix file system mount.
>
> Cheers, Steve.
>
>
>
> On 26 Oct 2007, at 22:16, Brad Goodyear wrote:
>
>> Well, at least, the commands associated with opening the file and
>> reading it doesn't work.
>> That is, the calls to fslio or niftiio fail, and say it cannot open
>> the file.
>> The same data on a local disk works just fine.
>> Has anyone else tried FSL on an Intel Mac with data on an SMB
>> mounted disk?
>> If anyone has any suggestion on how to make FSL happy with SMB
>> mounted disks, it would be greatly appreciated.
>>
>> -Brad
>
>
> ----------------------------------------------------------------------
> --
> ---
> 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, 27 Oct 2007 10:48:57 +0100
> From: Steve Smith <[log in to unmask]>
> Subject: Re: siena, sienax combined with betpremask
>
> Hi,
>
> If you want more complex BET calls (e.g. standard_space_roi as well
> as BET robust etc) then you can just take a local copy of $FSLDIR/bin/
> siena and edit the two BET lines near the top, putting in whatever
> you need. Don't forget though to make sure you generate the skull
> images as well as the brain and brain mask images.
>
> Cheers, Steve.
>
>
> On 26 Oct 2007, at 00:26, Antonios - Constantine Thanellas wrote:
>
>> Hi Steve,
>>
>> I've tried the -B option inside siena ( -B "-B") though in this case
>> automatically i exclude bet_robust choice from siena. There's no
>> way to feed
>> siena with the image that comes from betrobust segmentation of the
>> betpremasked image? I am asking that because when i applied
>> betrobust on an
>> image which was first betpremasked i had excellent brain segmentation
>> results and i hope that the same would happen if i could manage to
>> combine
>> both of the tools in siena
>>
>> When i used -B "-B" option on siena even though the bet results
>> were decent
>> (but definitely not as good as the results that i had using
>> betrobust on the
>> betpremasked image) and i thought that my problem was solved, i
>> noticed that
>> the halfway tranformations on both images were totally wrong
>> containing all
>> the neck... why this happened since bet extraction was done
>> successfully in
>> siena?What do you suggest to do in this case?
>>
>> I've uploaded the png files that shows the different stages of
>> siena (which
>> failed) on the A and B images with the ref number: 625404
>>
>> Thanks again Steve
>> Antonios-Constantine Thanellas
>>
>>
>>
>> On Thu, 25 Oct 2007 07:42:33 +0100, Steve Smith
>> <[log in to unmask]> wrote:
>>
>>> Have you tried the -B option to bet? You can pass that into SIENA by
>>> adding onto the end of the SIENA command
>>> -B "-B"
>>> Cheers.
>>>
>>>
>>> On 24 Oct 2007, at 20:52, Antonios - Constantine Thanellas wrote:
>>>
>>>> Dear fsl users
>>>>
>>>> I want to betpremask (through standard_space_roi) my data in order
>>>> to do
>>>> longitudinal and cross sectional analysis with siena and sienax
>>>> respectively.
>>>>
>>>> Since siena needs the skulls of the original images i can't just do
>>>> a shell
>>>> script that sends as inputs to siena the stripped brain images
>>>> after
>>>> betpremasking them. I've read in FSL forum that it's better to
>>>> change the
>>>> siena script as follows: "leave BET to extract the skulls, then run
>>>> betpremask and feed, then run again bet on the premasked images and
>>>> feed
>>>> siena with these results." The problem is that siena script is too
>>>> complicated for my scripting skills and i would really appreciate
>>>> if you
>>>> could tell me how to change this script in order to betpremask my
>>>> data first
>>>> (bet_robust totally fails in most of my data which contain big
>>>> protions of
>>>> neck even if i determine the COG with the -c option.That's why i
>>>> need
>>>> betpremask since it gives me really good results).I use the fsl 4.0
>>>> version
>>>> on ubuntu platform.
>>>>
>>>> How about sienax ? it also needs the skulls in order to register
>>>> the images
>>>> with pairreg? How the sienax script should be changed in order to
>>>> incorporate betpremasking?In this case also, a shell script that
>>>> sends as
>>>> inputs to sienax the betpremasked images wouldn't work..Right?
>>>>
>>>> Thank's a lot for your time
>>>>
>>>> Antonios-Constantine Thanellas
>>>
>>>
>>> --------------------------------------------------------------------
>>> -
>>> ---
>>> ---
>>> 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: Sat, 27 Oct 2007 10:53:30 +0100
> From: Steve Smith <[log in to unmask]>
> Subject: Re: can I implement the following in FEAT
>
> Hi,
>
> 1) It doesn't matter whether you call this block or event-related -
> the timings are the timings - just set them up with either of the
> "custom" options and the result will be identical.
>
> 2) Ah, no, I hadn't realised that your scanning was not continuous.
> In this case, no, things get a lot more complicated as you will have
> nasty T1 saturation effects to model.....see previous postings on the
> list by people like Andreas Bartsch and Joe Devlin on these issues.
>
> Cheers, Steve.
>
>
>
> On 26 Oct 2007, at 06:04, Xu Chen wrote:
>
>> Thanks, Steve.
>>
>> I am bringing up two more questions based on your answer .
>>
>> 1) are you suggesting me to analyze this data as block design
>> instead of event-related? I'd rather treat it as event-related,
>> but how?
>>
>> 2) As you may have noticed that the data collection is not
>> continuous in the sequence diagram. Only in steps 2, 5, 8,
>> 11......the data were collected. How can I deal with this in FSL?
>>
>> Thanks again
>>
>> Chen
>>
>> Steve Smith wrote:
>>> Hi - sounds like you have 3 conditions, including stim1=rest.
>>>
>>> Hence you should model this with just two EVs, one for when stim2
>>> is on and one for when stim3 is on. You can use either custom1 or
>>> custom3 to generate these.
>>>
>>> Cheers, Steve.
>>>
>>>
>>> On 25 Oct 2007, at 04:01, Xu Chen wrote:
>>>
>>>> Dear FSLers
>>>>
>>>> I am trying to analyze a fMRI dataset collected using the
>>>> following sequence diagram,
>>>> 1, 'stim1' ( on for 8 seconds)
>>>> 2, collect 4 scans (S11, S12, S13, S14) with 'stim1' off but
>>>> 'stim2' on (TR=2s) 3, rest for 8 seconds
>>>>
>>>> 4, 'stim3' ( on for 8 seconds)
>>>>
>>>> 5, collect 4 scans (S21, S22, S23, S24) with 'stim3' off but
>>>> 'stim2' on (TR=2s)
>>>> 6, rest for 8 seconds 7, 'stim1' ( on for 8 seconds)
>>>>
>>>> 8, collect 4 scans (S31, S32, S33, S34) with 'stim1' off but
>>>> 'stim2'on (TR=2s)
>>>>
>>>> 9, rest for 8 seconds
>>>>
>>>> 10, 'stim3' ( on for 8 seconds)
>>>>
>>>> 11, collect 4 scans (S41, S42, S43, S44) with 'stim3' off but
>>>> 'stim2'on (TR=2s)
>>>> 12, rest for 8 seconds .......................
>>>>
>>>> It is essentially an modified event-related experimental design.
>>>> 'stim2' ( actually 'the gradient noise') is always on in data
>>>> collection
>>>> steps only (step 2 , 5, 8, 11 ..... in above). And I want to apply
>>>> paired-t test on this data to get the contrast between 'stim1' and
>>>> 'stim3', namely, compare the pairs of (S11,S21) (S12, S22) (S13,
>>>> S23)
>>>> (S14, S24) (S31, S41) (S32, S42) (S33, S43) (S34, S44)..........
>>>>
>>>> I am not sure whether FEAT is capable to do this kind of
>>>> analysis. If
>>>> positive, how?
>>>>
>>>> Thanks
>>>>
>>>> Chen
>>>
>>>
>>> --------------------------------------------------------------------
>>> -
>>> ------
>>> 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: Sat, 27 Oct 2007 10:56:34 +0100
> From: Steve Smith <[log in to unmask]>
> Subject: Re: error in dtifit
>
> Hi - if you look at all timepoints of all slices of the data output
> by eddy_correct do they look ok? I.e., is the input to dtifit ok?
>
> Cheers.
>
>
> On 26 Oct 2007, at 17:39, Swati Rane wrote:
>
>> Hello,
>>
>> I use the older version of FSL (3.3). When I use the dtifit command
>> for my 3
>> slice data, the display says that it has processed all 3 slices,
>> but when I
>> look at the FA map in FSL view, I can see only the last 2 slices.
>> Can you
>> tell me what is the error in my processing?
>>
>> Thanks!
>>
>> Swati
>
>
> ----------------------------------------------------------------------
> --
> ---
> 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, 27 Oct 2007 11:02:17 +0100
> From: Steve Smith <[log in to unmask]>
> Subject: Re: Higher level analysis
>
> Hi,
>
> At the 3rd level, you should either use FLAME1 or FLAME1+2.
>
> Note that the FLAME options are not 'more stringent' than OLS, they
> are more accurate. (Both the FLAME and ME OLS options are the same
> basic model, i.e. mixed-effects.)
>
> Cheers.
>
>
> On 26 Oct 2007, at 14:57, rajani sebastian wrote:
>
>> Hi all
>>
>> I have a question regarding higher level analysis using FSL. My
>> fmri experiment is a blocked design and consists of two sessions
>> (approximately 6
>> minutes each session). I collected data from 10 subjects. I used
>> first level analysis for analyzing each sessions data. Then, i
>> used higher level analysis for combining the two sessions for each
>> subject using FIXED effects, cluster thresholding p< 0.05.
>> Finally, i plan to combine all the subjects together. The question
>> i have is regarding the stats section in higher level analysis.
>> Should i used simple OLS or FLAME 1+2? I read in the FSL manual
>> that simple OLS is the least accurate of all the ME options. So do
>> i use FLAME 1+2 as this is a suggested option if the number of
>> subjects is less. What i am concerned about is whether i would find
>> any activation since this procedure is very stringent. Also, what
>> thresholding level should i use, cluster or uncorrected or voxel
>> based ?
>>
>> Thanks
>> Rajani
>
>
> ----------------------------------------------------------------------
> --
> ---
> 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, 27 Oct 2007 11:08:10 +0100
> From: Steve Smith <[log in to unmask]>
> Subject: Re: Using SIENA in monkey MRIs
>
> Hi - it looks like things are nearly working. However, the alignment
> between the two timepoints are not very good, probably because there
> is very different bias field in the two images. I suspect that to get
> a better registration you will need to try using FAST to remove the
> bias before running SIENA. You can run FAST on brain-extracted
> images, with the bias output turned on and lots of bias field
> dilation. You can then apply this to the original data and hopefully
> remove the bias before running SIENA.
>
> Cheers.
>
>
>
> On 24 Oct 2007, at 01:54, Buyean Lee wrote:
>
>> Hi Steve,
>>
>> The six-digit reference number for this upload session is 754982.
>>
>> There are three compressed files.
>>
>> 1. whole SIENA output directory
>> 2. MRI before treatment (meanC9726_s1_2.img)
>> 3. MRI after treatment (meanC9726_s2_2.img)
>>
>> The two MRIs are the average images of the 9 Analyze format files
>> which were converted from the DICOM files (I acquired the same MRI
>> nine times in one session).
>>
>> Thanks a million for willing to take a look at the images.
>>
>> Buyean
>>
>> -----Original Message-----
>> From: Steve Smith <[log in to unmask]>
>> To: [log in to unmask]
>> Sent: Mon, 15 Oct 2007 11:04 pm
>> Subject: Re: [FSL] Using SIENA in monkey MRIs
>>
>> 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 <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.
>>
>> ---------------------------------------------------------------------
>> -
>> -----
>> 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: Sat, 27 Oct 2007 10:44:39 +0100
> From: Steve Smith <[log in to unmask]>
> Subject: Re: mask.nii.gz problem in 3rd level feat
>
> Hi,
>
> On 25 Oct 2007, at 14:45, David V. Smith wrote:
>
>> Hi-
>>
>> I'm using FSL 4.0.1, and I'm having some masking problems with FEAT
>> once I get
>> to the 3rd level analyses. The mask.nii.gz image created by FEAT
>> masks out
>> much more of the brain than I would like it to... It's essentially
>> cutting off
>> many parts of the brain and my stat maps look 'speckled' because of
>> the mask
>> (e.g., a big blob of activation can have a big chunk missing where
>> you might
>> expect to see a local maximum). This is particularly problematic
>> in FFA and
>> OFC. See attached images of 3rdlevel mask.nii.gz (shown in red)
>> overlayed onto
>> the 3rdlevel mean_func.nii.gz and also on the bg_image.nii.gz.
>>
>> 1.) Is this because these particular voxels are missing in some
>> subjects (and
>> shouldn't be used in the analysis) because of BET being a little
>> too liberal at
>> earlier stages in the analysis? I realized it's supposed to fairly
>> conservative
>> on the functional data, but clearly parts of my brains are getting
>> cut off...
>
> Yes. When you ran the second-level analysis there would be a
> registration evaluation summary page, which you should look at
> carefully, including all the registrations and the summary mask
> images at the bottom. You should definitely check which of the
> subjects were problematic.
>
> You're right that the thresholding and BET settings at first level
> should be overinclusive, so we don't often see this kind of problem.
> If your data has some weird intensity range, for example, you might
> need to change the %threshold setting (in the Misc tab) in the first-
> level analyses.
>
> This may fix things, also, if BET really isn't working well (unlikely
> for FMRI data) then you can just turn it off in the Pre-stats tab.
>
> As a last resort, yes you can over-ride the masking with a non-GUI
> option - copy
> mkdir -p ~/.fslconf
> cp $FSLDIR/etc/fslconf/feat.tcl ~/.fslconf
>
> and then edit ~/.fslconf/feat.tcl, changing the following line to
> include a full pathname to the mask you want:
> set fmri(alternative_mask) ""
>
> Note that this probably only has effect at the first-level analyses -
> there's nothing that you can do to "fix" the masks at higher-level -
> there's no point trying to fix the masks as the input data (copes
> etc) are already masked at first level.
>
> Cheers.
>
>
>> 2.) If it is related to the BET settings that FEAT/MELODIC use
>> during pre-stats,
>> how can I change these to something more appropriate (e.g., reduce
>> the
>> fractional intensity option)? If this is the fix, would I need to
>> re-run my
>> entire analysis?
>>
>> 3.) Is there a way to simply give FEAT a different mask instead of
>> the one it
>> generates? This seems to be the easiest fix and I thought one of
>> the new
>> non-GUI options (below) in the higher-level design.fsf would help
>> me, but it
>> didn't get used as the mask.nii.gz image like I expected. Is this
>> not what
>> this option is for?
>>
>> # Alternative (to BETting) mask image
>> set fmri(alternative_mask) "insert_better_mask_here"
>>
>> Thanks,
>> David
>> <bg_image.nii.png>
>> <mean_func.nii.png>
>
>
> ----------------------------------------------------------------------
> --
> ---
> 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, 27 Oct 2007 10:28:18 -0500
> From: rajani sebastian <[log in to unmask]>
> Subject: Re: Higher level analysis
>
> ------=_Part_12695_1804524.1193498898030
> Content-Type: text/plain; charset=ISO-8859-1
> Content-Transfer-Encoding: 7bit
> Content-Disposition: inline
>
> thank you.
>
> On 10/27/07, Steve Smith <[log in to unmask]> wrote:
>>
>> Hi,
>>
>> At the 3rd level, you should either use FLAME1 or FLAME1+2.
>>
>> Note that the FLAME options are not 'more stringent' than OLS, they
>> are more accurate. (Both the FLAME and ME OLS options are the same
>> basic model, i.e. mixed-effects.)
>>
>> Cheers.
>>
>>
>> On 26 Oct 2007, at 14:57, rajani sebastian wrote:
>>
>>> Hi all
>>>
>>> I have a question regarding higher level analysis using FSL. My
>>> fmri experiment is a blocked design and consists of two sessions
>>> (approximately 6
>>> minutes each session). I collected data from 10 subjects. I used
>>> first level analysis for analyzing each sessions data. Then, i
>>> used higher level analysis for combining the two sessions for each
>>> subject using FIXED effects, cluster thresholding p< 0.05.
>>> Finally, i plan to combine all the subjects together. The question
>>> i have is regarding the stats section in higher level analysis.
>>> Should i used simple OLS or FLAME 1+2? I read in the FSL manual
>>> that simple OLS is the least accurate of all the ME options. So do
>>> i use FLAME 1+2 as this is a suggested option if the number of
>>> subjects is less. What i am concerned about is whether i would find
>>> any activation since this procedure is very stringent. Also, what
>>> thresholding level should i use, cluster or uncorrected or voxel
>>> based ?
>>>
>>> Thanks
>>> Rajani
>>
>>
>> ---------------------------------------------------------------------
>> ---
>> ---
>> 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
>> ---------------------------------------------------------------------
>> ---
>> ---
>>
>
> ------=_Part_12695_1804524.1193498898030
> Content-Type: text/html; charset=ISO-8859-1
> Content-Transfer-Encoding: 7bit
> Content-Disposition: inline
>
> thank you.<br><br><div><span class="gmail_quote">On 10/27/07, <b
> class="gmail_sendername">Steve Smith</b> <<a
> href="mailto:[log in to unmask]">[log in to unmask]</a>>
> wrote:</span><blockquote class="gmail_quote" style="border-left:
> 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-
> left: 1ex;">
> Hi,<br><br>At the 3rd level, you should either use FLAME1 or FLAME1
> +2.<br><br>Note that the FLAME options are not 'more
> stringent' than OLS, they<br>are more accurate. (Both the FLAME
> and ME OLS options are the same
> <br>basic model, i.e. mixed-effects.)<br><br>Cheers.<br><br><br>On
> 26 Oct 2007, at 14:57, rajani sebastian wrote:<br><br>> Hi
> all<br>><br>> I have a question regarding higher level
> analysis using FSL. My<br>> fmri experiment is a
> blocked design and consists of two sessions
> <br>> (approximately 6<br>> minutes each session). I
> collected data from 10 subjects. I used<br>> first
> level analysis for analyzing each sessions data. Then,
> i<br>> used higher level analysis for combining the two sessions
> for each
> <br>> subject using FIXED effects, cluster
> thresholding p< 0.05.<br>> Finally, i plan to combine all the
> subjects together. The question<br>> i have is regarding the
> stats section in higher level analysis.<br>
> > Should i used simple OLS or FLAME 1+2? I read in the FSL
> manual<br>> that simple OLS is the least accurate of
> all the ME options. So do<br>> i use FLAME 1+2 as this is a
> suggested option if the number of<br>> subjects is less. What i
> am concerned about is whether i would find
> <br>> any activation since this procedure is very stringent.
> Also, what<br>> thresholding level should i use, cluster or
> uncorrected or voxel<br>> based ?<br>><br>>
> Thanks<br>>
> Rajani<br><br><br>----------------------------------------------------
> --------------------
> <br>---<br>Stephen M. Smith, Professor of Biomedical
> Engineering<br>Associate Director, Oxford University
> FMRIB Centre<br><br>FMRIB, JR Hospital, Headington,
> Oxford OX3 9DU, UK<br>+44 (0) 1865 222726
> (fax 222717)<br><a href="mailto:[log in to unmask]">
> [log in to unmask]</a> <a href="http://
> www.fmrib.ox.ac.uk/~steve">http://www.fmrib.ox.ac.uk/~steve</
> a><br>----------------------------------------------------------------
> --------<br>---<br></blockquote></div><br>
>
> ------=_Part_12695_1804524.1193498898030--
>
> ------------------------------
>
> Date: Sat, 27 Oct 2007 21:33:09 +0100
> From: Ying Chen <[log in to unmask]>
> Subject: Re: Skull and scalp extraction
>
> Hi,
>
> I used T1 and T2 image set provided in FSL example data, and I did
> get=20=
>
> four separate voxel maps which correspond to scalp, skull, brain
> and the=20=
>
> region between brain and skull, respectively.=20
>
> Does the region between brain and skull represent cerebro-spinal
> fluid?=20=
>
> But the brain segmentation also includes an outer layer of CSF. So
> I am=20=
>
> not sure what part of human head one of the voxel maps I am getting
> (the=20=
>
> one between brain and skull) represents.=20
>
> Thank you!
>
> Ying
>
> ------------------------------
>
> End of FSL Digest - 26 Oct 2007 to 27 Oct 2007 (#2007-293)
> **********************************************************
|