Hello,
We are getting some inaccurate freesurfer segmentations on our T1w pediatric data. Running synthseg on the same data improves the segmentation results significantly, but I was wondering how to use these improved segmentations in fmriprep pipeline. I thought of two options and would appreciate any input if anyone has experience trying to do this:
Option 1- Use synthseg segmentations to edit the segmentation from recon-all. Then run fmriprep and point to that editted freesurfer directory. It is unclear to me what would be the steps needed to feed the synthseg segmentation into recon-all to generate the surfaces, since synthseg only outputs a single nifi file. Is there a way to run specific steps of recon-all? could not find clear documentation.
Option 2- run recon-all-clinical (recon-all-clinical - Free Surfer Wiki) instead of recon-all, then point fmriprep to that folder. When I tried this, fmriprep crashed, since the freesurfer/sub-xx/mri/orig folder generated by recon-all-clinical was empty. Can I copy the contents of the orig folder from a separate recon-all run? The issue could also be related to the fact that fmriprep contains freesurfer 7.3.2, while recon-all-clinical is only available from freesurfer version 7.4, was wondering if there’s a compatibility issues?
I’d appreciate any leads or ideas!
Thanks
Maya