Il giorno 30 lug 2023, alle ore 15:32, Michelangelo Fabbrizzi <michefabb@gmail.com> ha scritto:

Dear Freesurfer experts,

I am new to FS, and I would have couple of questions.
I used in sequence:
A- recon-all -i sub-01_T1w.nii.gz -s sub-01_recon -autorecon1
B- recon-all -skullstrip -wsthresh 5 -clean-bm -s sub-01_recon
C- recon-all -skullstrip -clean-bm -gcut -subjid sub-01_recon

The brainmask.mgz obtained is good (no skull, brain is okay), however if I now use mri_watershed to obtain the bem surfaces, it looks the brainmask.mgz is not used and it is created a new one called ws.mgz that still contain some piece of skull that generate wrong BEM surfaces; even using -h 5 the result is not optimal:
mri_watershed -h 5 -useSRAS -surf /Users/michelangelo/Test_skullstripping/sub-01_recon/bem/watershed/sub-01_recon /Users/michelangelo/Test_skullstripping/sub-01_recon/mri/T1.mgz /Users/michelangelo/Test_skullstripping/sub-01_recon/bem/watershed/ws.mgz

I am wondering:
- how can I force the mri_watershed command to use the brainmask.mgz already created after step C above, instead of creating a new one?
- is it possible to use the -gcut command also for mri_watershed?

Thanks, appreciated,

Miche