-atlas checks the deformation field against an atlas to see if it's reasonable. If it not (e.g. if the cerebellum is stripped), it tries to fix it. As for BET or mri_watershed, it's completely empirical. Whichever works better for you is fine.
cheers, Bruce
On Thu, 2 Jun 2005, Fornito, Alexander wrote:
Hi all, The page about using FSL with freesurfer on the FSL website recommends using BET to skull strip the images before feeding them into freeusrfer. I've had a bit of play with this, and have noticed some marginal improvements. Would you generally recommed this practice? Are there any reasons not to? Also, what exactly does the -atlas option do? Thanks, Alex
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer