Dear Experts, 

I have read the previous posts in the list regarding restricted ROI analysis. So I generated my label from tksurfer and used that for group difference analysis by adding --label to mri_glmfit, but the result (sig.mgh) seems just a masked version of the analysis including whole brain ( no difference in significance! ). I expected to see a different pattern of significance by limiting the region. Would you please advise if I am doing this in a right way?