Dear Alfonso,

Thank you in advance for your help.

1. I would like to analyse ICA networks in all subjects so, if I am not wrong, I should go to voxel-to-voxel measures, select the contrast [1] for allsubjects and select all ICA and I will obtain my results. The image of the brain is almost completely red (which it makes sense because all areas have some degree of connectivity). However, when I adjust them for age, gender and lesions, the image doesn't show anything in red and I don't obtain any significant cluster. even when I choose a really low conservative threshold. If I am not wrong, this can't be possible, right? Why could this happen? I am using the contrast [1 0 0 0 ] (where 1 is allsubjects and 0 are the three covariates) for allsubjects and when I want to compare two groups (controls>patients) I use the contrast [-1 1 0 0 0] (where -1 is controls, 1 is patients and the three 0 are the three covariates). Is this right?

2.I would like to compare functional connectivity between two groups. I would be very grateful if you could tell me if it would make sense to do a simple t-test between the two groups (controls>patients) using the contrast [-1 1] and selecting all ICA (eye 20). I doubt whether ICA's are a descriptive measure of functional connectivity and I should use other measures in order to see differences between groups (I have read some papers in which they do ICA measures and then compare all ROI's belonging to each ICA between groups, for example, or others do ICA and then dynamic FC measures or graph theory).



Many thanks,

Marta