Dear FSL experts
I would like to explore differences of the sensorimotor network between patients (N=20) and healthy (N=20). I have ran Melodic (with autodim) and recognized the component of interest (sensorimotor network) from the output of circa 120 components.
When I perform dual regression it takes ages to complete because the group-comparisons (randomise) are done for all of the 120 components.
1. How can I perform the randomise-step only for my component of interest?
2. I would like to explore differences between a subgroup of patients (n=10) and healthy (either n=10 or N=20). Should I proceed a) with running a Melodic on the subgroup etc., or b) can I use the components from the whole-group Melodic, and if I can, how can I setup the subgroup comparison?
Best regards
Jaakko
|