I don't think the brain.mgz is the right one to look at as it has already been partially segmented. Compare against the norm.mgz (or nu.mgz) instead. It is really hard to determine where the lateral boundaries are in thalamus
On 8/31/2022 2:46 AM, Thalhammer, Melissa wrote:
External Email - Use Caution
Dear Freesurfers,
I have used the Freesurfer-based thalamus subnuclei segmentation. As a quality control, we have inspected the output files visually in freeview and noticed that some (roughly 5 out of 25 subjects) thalamic volumes are (much) larger than the brain.mgz file would indicate (see attached screenshot showing brain.mgz and ThalamicNuclei.v12.T1.FSvoxelSpace.mgz). This seems to be the case for aseg.mgz as well, i.e., the thalamus in aseg.mgz seems larger than it should be based on the contrast we see in brain.mgz.
We have tried to run the segmentation again but this has not led to significantly better results. Is this the right way to visualize the outputs? Is it recommended to exclude these subjects? Is there any recommended quality control pipeline that we could apply?
Thank you a lot in advance! Best, Melissa Thalhammer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer