Hi all -- I'm still having difficulty with 2 problems that I think are solvable, but I can't figure them out.
1) When I call up aparc+aseg on the T1 image, I get the cortex that has been parcellated, but I do not get the parcellation -- that information is in the .annot files that seem to be viewable only using tksurfer and not tkmedit. Is there a way to see the full parcellation on T1.mgz?
2) Trying to run freesurfer on only the Right hemisphere on a child who has no left hemisphere, I've run into the problem where freesurfer wants to divide the intact hemisphere into half. When I try to use the seed points, it hasn't fixed the problem. Is this a problem of syntax or something else?
These are the commands I used that I thought would work:
mri_fill -C -93 -18 89 -P -35 -22 28 -rh -59 -23 105 subjects/005/mri.mgz subjects/005/mri/filled.mgz
That part seemed to work, then:
recon-all -hemi rh -s 005 -autorecon2-wm
This part seemed to work until it got to CORRECTING DEFECT 1, where it said recon-all exited with ERRORS after it said "reading wm segmentation from wm.."
Have I missed a step or misunderstood how to do the seed points?
Thanks again, Joan
Hi Joan,
- When I call up aparc+aseg on the T1 image, I get the cortex that has been
parcellated, but I do not get the parcellation -- that information is in the .annot files that seem to be viewable only using tksurfer and not tkmedit. Is there a way to see the full parcellation on T1.mgz?
The .annot files can only be loaded on top of a surface in tksurfer. As you saw, the aparc+aseg.mgz is a segmentation that can be viewed in tkmedit, over the T1.mgx volume. What information were you hoping to get from the .annot file loaded on the T1.mgz?
- Trying to run freesurfer on only the Right hemisphere on a child who has
no left hemisphere, I've run into the problem where freesurfer wants to divide the intact hemisphere into half. When I try to use the seed points, it hasn't fixed the problem. Is this a problem of syntax or something else? These are the commands I used that I thought would work:
mri_fill -C -93 -18 89 -P -35 -22 28 -rh -59 -23 105 subjects/005/mri.mgz subjects/005/mri/filled.mgz
That part seemed to work, then:
recon-all -hemi rh -s 005 -autorecon2-wm
-autorecon2-wm starts with the mri_fill step, effectively re-doing what you had done in your own command, but without the extra options. Try running your mri_fill command again, and if the resulting filled.mgz volume looks ok, then run: recon-all -hemi rh -s 005 -autorecon2-wm -nofill which will use your filled.mgz volume from the mri_fill step and run tessellate onward. If you are using a very recent dev release there is a flag to start at this point: recon-all -hemi rh -s 005 -autorecon2-perhemi
Jenni
freesurfer@nmr.mgh.harvard.edu
-
Jenni Pacheco -
Joan Fisher