Hello,

 

We are using FreeSurfer v5.3. and are sometimes finding that a brain can seem okay when it is first run through the recon-all process and is edited, but then when it is re-run the segmentation seems to deteriorate hugely. I have screenshotted an example below. 

 

Are you able to advise us on why this might be happening and what we could do about this? One of the options I had thought about was to re-run the recon-all from scratch but it doesn’t seem very time efficient to do this for multiple brains. Any suggestions would be gratefully received!

 

Thank you.