Hi,
This bit of code is actually not for general analyses - it was written
for a very specific task and does only calculate values at the edge
of the mask. We don't recommend using it for the kind of analysis
that you are doing.
Instead, you should use the GLM gui and fsl_glm to do this
analysis. It is easy to set-up the group difference you want with
the Wizard in the GLM gui, and once you've saved your design
then you can run fsl_glm just using the -i, -o, -m and -d options.
You will also need to run fslmerge to put all of your FA images
into one 4D file (make sure you have one group followed by the
other group and not interspersed).
All the best,
Mark
On 31 Dec 2007, at 03:51, Matthew Hoptman wrote:
> Hi all,
> I have been trying to examine group differences in FA using
> groupttest. I
> am using a white matter mask to limit my analysis to that tissue
> compartment. The command I used is:
>
> groupttest --na=20 --nb=20 -m wmmask -o groupdiff <group a images>
> <group
> b images>
>
> I noticed that when I run the program, it only seems to test the
> voxels on
> the edge of the mask (p-values, etc., are 0 inside the mask). Is
> this the
> expected behavior? If so, what would be the best way to test these
> data?
>
> Thanks and happy holidays,
> Matt
>
|