External Email - Use Caution
Hi all,
I'm encountering an issue with surface reconstruction quality in
ventromedial prefrontal cortex, using relatively high resolution (.8mm) T1-
and T2-weighted input data to recon-all. I'm using the following command,
with Freesurfer 7.1.1 running on Mac OS 10.15.7
(freesurfer-darwin-macOS-7.1.1-20200811-8b40551; the options file specifies
"mris_inflate -n 30"):
recon-all -s ${subject} -i T1w.nii.gz -all -hires -expert ${optsFile} -T2
T2w.nii.gz -T2pial
Across four participants I've tested with this approach, there is a
consistent reconstruction issue with the gyrus rectus and medial orbital
gyrus, in which they are cut off in the white and pial surfaces (see images
at
https://secure-web.cisco.com/1YjHwYMb6FHlCVkbgjtpceoj_IZMMa8YVArerusSJpfEDYCcAf_7XW4Eyr9GSgrttGniK959dynsL2QmzsmEonYXtiwqdr8sJaV6qRLe8n7426mcHuELvluFukkaEVt1rBzgplsCCLyRUFdccKTe3oEh8eHH1HOyOPmfSrG6dFq8ivGCTGju2_vmz4AsNFp8RmYuqWLBkZmqb7urX9xgtAtB2SfRXXM6eUm0ETRM2FYqgWafJFE2sR-EM1osxiFAN-HFcudJZYS4Jq5l4_nbGGw/https%3A%2F%2Fwww.dropbox.com%2Fsh%2Fcrpcjro63bc6477%2FAADNtt9xJL1SV2kiaVJsGnSYa%3Fdl%3D0%3B
it's particularly clear against the T2 image). I've tried to correct the
issue by manually modifying wm.mgz to include all wm voxels in these gyri.
However, when I do so and rerun autorecon2-wm, the orig surface improves to
incorporate the edits to wm.mgz, but the white/pial surfaces do not. By
debugging individual steps of recon-all, I've found that the -white step
seems to be to blame for this: when I rerun this step, the output
white.preaparc surface doesn't change to incorporate my edits to wm.mgz.
Any thoughts as to what might be causing these issues - both the initial
reconstruction problem, and the failure to update with manual edits? There
is some signal dropout in the T1 image in this region, which may contribute
to the issue, although I would hope the T2 image could compensate for this.
Thanks,
Ben
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
