Dear Freesurfers,
My goal is to evaluate the cortical thinning patterns in some (cortical aparc) regions, and to relate that to loss of PET activation in the same regions. I would like to clarify the steps of the pipeline in Freesurfer (v5.0.0) for this purpose.
From my reading of previous discussions in the mailing list, I can see
two possible ways of doing this - but I am not sure and I would like to ask for your feedback on them:
First method: 1) Register all PET scans (already in ICBM152 space, enforced with SPM8) to the brain.mgz image of each subject to ensure the PET files are in MNI305 space, to be able to use Freesurfer's atlas, using bbregister or spmregister. 2) Use mri_vol2surf to project the PET data onto the surface of each subject. 3) Use mri_segstats to extract the statistics
for each region.
Second method: 1) (Same as above). 2) Use mri_annot2label on the relevant aparc regions. 3) Use results from the previous step in mri_label2vol to create the volume masks extracted from the cortical regions. 4) Apply mri_segstats on these (as above).
Could you please clarify if I am on the right track, notably which method should be more robust (or correct...)? Also, does mri_segstats also give overall group statistics (mean, std)
for each region, or must I use another function/average all subjects together for that?
Thank you very much!
Best,
Joao Pereira Wolfson Brain Imaging Centre University of Cambridge