These are voxels that are labeled as cortex but do not have a nearby annotation in the parcellation, so I just labeled them as "unknown". I don't think there is an easy way to relabel them as something else in v6. This is fixed in version 7. You can run the v7 program on the v6 data. If you want to do this, download v7. Make a copy of one of your subjects, then run

recon-all -s copy.subject -aparc2aseg

This will create new aparc+aseg.mgz using the new program. You can then get the mri_surf2volseg command line and apply it to all of your subjects (changing the name of the output as you like). You could just run the v7 recon-all command above on your v6 data, but it will change things like the build stamp and the log file, which might be confusing when trying to keep track of what was done to the data.




On 6/16/2020 4:21 PM, Caspar M. Schwiedrzik wrote:

        External Email - Use Caution        

Hi!
We are facing some issues in the aparc+aseg of the medial temporal lobe.
As you can see on the image, there are quite a few voxels labeled as unknown in both hemispheres, and generally, there seems to be a bit of a mess in that general area in the aparc+aseg (above), while everything is marked as cortex in aseg (below). We see this pattern in a few subjects. Is there a way to get these voxels assigned to a structure? wm edits have not fixed the issue so far. Thank you!
Caspar

aparc.png

_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer