sorry, not easily. I guess you could recon them in the newer version then try using mri_label2label to transport them to the older one, but it isn't trivial since the fsaverage coordinate system has changed
Bruce
On Mon, 10 Apr 2017, Taylor, Johnmark wrote:
(the reason I did this was because the new Freesurfer version outputs labels for perirhinal cortex as part of the reconstruction process, and we were interested in using this ROI in our analyses; as an additional question, is there a way of producing these anatomical labels without reconstructing the surface from scratch?)
On Mon, Apr 10, 2017 at 5:46 PM, Taylor, Johnmark johnmarktaylor@g.harvard.edu wrote: Hello, Recently, I re-ran recon-all with Freesurfer version 5 on some subjects whose surfaces had previously been reconstructed with Freesurfer version 4. As a result, it appears that their surface reconstructions have subtly changed, such that the labels that were specified in the old reconstruction are all messed up in the new reconstruction (since the surfaces that they were once associated with have changed). Did recon-all change between Freesurfer versions, and if so, is there a way to go back to the old algorithm short of simply switching back to Freesurfer version 4? It is possible that I have accidentally changed something else and that's why the surface files are different, but I wanted to see if there were any changes in the reconstruction algorithm between versions that might have caused this.
Thank you very much,
JohnMark Taylor