Print

Print
Hi Andrew,

On 18 September 2015 at 16:53, Andrew Song <[log in to unmask]> wrote:
Dear FSL experts,

I just wanted to run by whether my approach in group analysis is logically sound.

To incorporate CompCor (using WM/CSF principal components as the regressors in GLM), I chose to do individual first-level GLM in each individual's native (or structural) space, not the standard MNI space.

After having obtained COPE/VARCOPE files for each individual, I apply FLIRT/FNIRT transformations on these COPE/VARCOPE files, so that I can bring all these data into the MNI space for group-level analysis.

If applying transformation on COPE/VARCOPE files is not feasible, what are the possible alternatives?



If you've done standard first-level FEAT GLM analyses with registration, a group level FEAT analysis will automatically convert the copes/varcopes into standard space as part of the analysis. Is this your scenario?

Cheers

Eugene

 
Thanks,