Hi,

 
We have run recon-all on 33 healthy subjects in their 20ies with FreeSurfer version 5.3.0, on a Linux CentOS 6.7. The original T1's look good 
(example image attached). We used the braindeer tool(http://brainder.org/2011/09/10/quickly-inspect-freesurfer-cortical-surfaces/) to check the quality
 of the cortical reconstruction, and we find that in 10-15 subjects the cortical surface reconstruction was unsuccessful. We think it may 
have to do with an incomplete skull strip (please see attached images). In order to get rid of residual dura (and some skull) we changed 
watershed first to 20 then to 5, using these commands:
 
1) recon-all -skullstrip -wsthresh 20 -clean-bm -subjid <subject name>
2) recon-all -autorecon-pial -subjid <subject name>

( We also tried: recon-all -skullstrip -clean-bm -multistrip -subjid <subject name>  )
 
None of this changed the outcome much. Neither did -gcut .....

An alternative is of course manual editing, but we are wondering whether using BEaST to do the skull strip, and then replacing the brainmask.mgz 
with the brainmasking resulting from BEaST (and then recon-all -autorecon-pial -subjid <subject name> ), could be a workable solution.

Would you recommend to do that/ not to do that?
 

Thank you!






sincerely yours,

Lars M. Rimol, PhD
Senior researcher,
Norwegian Advisory Unit for functional MRI
Department of Radiology,
St. Olav's University hospital,
7006 Trondheim,
Norway