Hi Steve, sorry I wasn't very clear. I'm interested in comparing the effect of two different tasks. I have a group of people that performed both. However, if I understand it correctly I can't concatenate those images as they have different number of volumes ( and the task is different). Therefore I run temporally concatenated group ICA for each task separately. Next, I run fslcc to identify closely correlated components between the two tasks. Next I wanted to do a more quantitative comparison, therefore I wanted to extract individual time series for each subject for those components (from "t#.txt" files) and use them as input for the FEAT analysis. That way I have FEAT for each subject for each component of interest (i.e. those closely correlated components between tasks). Using featquery I can then extract and do a paired comparison in the ROIs between two tasks (in closely correlated components).
Hope that explains it a bit better. My main concern was can I use the time series estimates produced from ICA as an input in the FEAT
I really appreciate your help and time
Best
Ekaterina
|