I have processed data using freesurfer and the output for most of my subjects did not remove all of the eye and dura around it during skull stripping. When I check the brainmask there is dura (or blood vessel) in the frontal lobe that is classified as GM.

I am wondering if I have to manually edit the skullstrip and rerun recon-all or do I just edit the brainmask file by deleting non-brain voxels? Or do I have to do both? 

If I only edit the brainmask file does this skip any other important processing steps that may affect my results?

I am planning on looking at brain volumes and cortical thickness and how they relate to behavioral measures.

Kind Regards,

Ciara