I have three groups of patients (healthy controls, patient group 1,
patient group 2) on whom I have done a TBSS analysis. I have performed
the standard comparisons FA control > group 1, FA control < group 1, FA
control > group 2, FA control < group 2 by removing either the group 1
or group 2 scans from all_FA_skeletonised and using design_ttest2 as per
the TBSS manual.
Now what I would like to do is to relate the FA measurements to
psychological variables by asking the question "Where in healthy
controls / patient group 1 / patient group 2 does FA correlate with
psychological variable X?" (as you can with a whole brain voxel-based
analysis in SPM). I have the values of the psychological variable as a
vector.
It looks like I might have to do something with Glm / Feat to create an
appropriate design matrix but cannot work out exactly what I need to do.
Obviously, I do not want to treat the psychological variable as a
nuisance variable. Any help would be appreciated.
Many thanks,
Gavin
|