Hello FreeSurfer Developers,

I have been attempting to compare FSL lesion-filled patients to their non lesion-filled counterparts. However, with this particular subject the skull-stripping has both over estimated and under estimated areas within the brain. I’m attaching a few pictures to show what I’m referring to. I have tried using both FSL (bet), as well as, performing watershed manipulation as described in the recon-all step -skullstrip, but both haven’t been able to fix the problem. I was wondering if you guys would know the best approach to solving this problem?

I have tried to search for similar skull-stripping discussions, but did not find any other possible solutions.

• FreeSurfer version: freesurfer-Darwin-OSX-stable-pub-v6.0.0-2beb96c
• Platform: macOS Mojave v10.14.5
• recon-all.log: see attached

Amy Paulson

Research Assistant

University of Minnesota

Department of Pediatrics | 717 Delaware

Minneapolis, MN 55414

Phone: 952-288-5110