I am having trouble getting a usable brainmask for one of my subjects - a screenshot of the initial brainmask.mgz is attached as image_1. The talairach transformation looks fine, so I tried running recon-all -multistrip with raised watershed heights to no avail - see image_2.
Does this mean the problem is with the mri_em_register step / lta output? How can I check this, and what can I try changing?
Thank you!