I have completed recon-all on my dataset of controls and patients. My patients have hippocampal/MTL atrophy and I would like to conduct multiple comparisons on the data to ensure that the atrophy is indeed focal in each patient and if not, identify where else atrophy is occurring in these patients. My hypothesis is that I do not expect atrophy elsewhere due to the disease type of these patients, but as this has never been confirmed outside of visual inspection before I wish to check this more robustly.

I am not at all familiar with this kind of analysis and would like to ask some advice on how I should approach this? My initial instinct is to go with permutation testing. I would like to consider the whole brain (subcortical and surface).