Hi Dan
that sounds worrisome. I don't think we've seen this problem in our test cases. Can you upload an example so we can take a look?
thanks Bruce On Fri, 22 Mar 2013, Daniel Carey wrote:
Dear All,
We have been running the recon-all pipeline in 5.2 and have been encountering some difficulties with segmentations. We have isolated these issues to problems with Talairach registrations, and have been re-running recon-all with the Talairach registrations manually corrected in tkregister2; however, doing so appears to worsen our segmentation problems. In particular, the orig surface appears substantially shrunk away from the filled and white matter volumes.
Has anyone encountered such difficulties with manually corrected Talairach registrations placed in the processing stream, or indeed, seen the surface shrink to such an extent? We initially provided recon-all with synthesised volumes obtained using MRI-synthesise from the developer version (our raw scans are high-resolution PDw and T1w volumes).
Any insights would be greatly appreciated.
Many thanks,
Dan
Daniel Carey, M.Sc., Marie Curie PhD Fellow, Centre for Brain and Cognitive Development, Birkbeck College, 32 Torrington Sq., London, WC1E 7JL UK
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer