Thanks Bruce. After I remove the problem voxels in wm.mgz, do I then re-run the recon from autorecon2-wm?
Thanks, Matt ________________________________________ From: freesurfer-bounces@nmr.mgh.harvard.edu [freesurfer-bounces@nmr.mgh.harvard.edu] on behalf of Bruce Fischl [fischl@nmr.mgh.harvard.edu] Sent: Friday, May 26, 2017 1:40 PM To: Freesurfer support list Subject: Re: [Freesurfer] 7T recon
Hi Matt
yes, that means there was a topological defect that was incorrectly fixed. If you load the wm.mgz over the lh.orig.nofix, then in the orig.nofix load the file defect_labels as an overlay and change the overlay settings to that the min is 1 and click the "use overlay color" button it will show you the regions of the surface that are defective. You will then need to edit the wm.mgz to remove the handle/hole that is the defect.
cheers Bruce
On Fri, 26 May 2017, Hibert, Matthew Louis wrote:
Hi All, I reconstructed 20 subjects with a 750um isotropic MPRAGE collected at 7T and while most look good, 4 of the subjects have a similar section of the left STG removed from the final surfaces (picture attached). In all 4 subjects the entire left STG is included in the wm.mgz and lh.orig.nofix (red), but a portion has been removed in lh.orig (yellow), so the problem seems to occur in the -fix stage of autorecon2.
How can I go about correcting this issue?
Thanks, Matt