The FSGD file is incorrect. It has multiple subjects listed, and the string after the subject name is not the name of a class (Tract1_FA is not the same as Tract1). I would suggest working through the example on the wiki tutorial and/or looking at the slides on group analysis. doug

On 06/12/2012 06:05 PM, Shantanu Ghosh wrote:

Hi Doug

I set up my .fsgd file as follows (4 tracts of interest, 59 subjects belonging to 2 different groups, CON and PAT):

GroupDescriptorFile 1 Title tract_fa MeasurementName thickness Class Tract1 Class Tract2 Class Tract3 Class Tract4 Input Subject1 Tract1_FA Input Subject1 Tract2_FA Input Subject1 Tract3_FA Input Subject1 Tract4_FA Input Subject2 Tract1_FA Input Subject2 Tract2_FA Input Subject3 Tract3_FA Input Subject3 Tract4_FA . . Input Subject59 Tract4_FA

and set up the .mtx file for mri_glmfit as:

C = [zeros(3) eye(4)]

Is that correct? How do I define the contrast variables = -1 and 0 for the test?

I am a little confused.

Thanks and regards, shantanu

On Tue, June 12, 2012 4:57 pm, Douglas N Greve wrote:

...

Did you look at the FSGD examples on our wiki? Do a search for FSGD doug

On 06/12/2012 04:46 PM, Shantanu Ghosh wrote:

...

Hello Freesurfer experts,

How can I specify multiple contrasts for my thickness study with two groups AA and BB, having different number of subjects in each group? My variables are tract-wise FA values, age and IQ.

Thanks in advance, Shantanu

-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422

Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html