1. Yes. 2. Ignore everything outside of the ROI and do a bonferroni correction by the # of vertices in the ROI.

cheers Bruce

On Thu, 24 Mar 2011, Dalwani, Manish wrote:

Hello Freesurfers,

We are interested in comparing cortical thickness in 2 groups. Due to small sample size (and thus less power), we are swaying away from whole-brain vertex analyses. Instead we would like to compare specific ROIs between two groups. We understand that we can get an estimate of average cortical thickness from each ROI. We fear that by comparing those estimates over a large ROI may cause it to be less sensitive in picking up a very small region of severe cortical thinning.

Therefore, our questions are as follows:

1. Can we perform vertex-by-vertex analyses between groups but not on whole-brain and only on the specific apriori ROIs?
2. If we can do vertex-by-vertex ROI analyses, is there a reasonable statistical method to correct for multiple comparisons?

Kindly advise!

Regards, Manish Dalwani Sr. PRA Dept. of Psychiatry University of Colorado Denver Tel: 303-724-3185 Email: manish.dalwani@ucdenver.edu