Re: [Freesurfer] Bad pial surface - continued

-- On Mon, 10 May 2010, Ed Gronenschild wrote: > Hi, > > I've mailed this once before and tried to solve a persistent problem. > Reviewing the results of recon-all I noticed that the derived pial > surface chopped away an unacceptable amount of gm voxels. > First of all I tried to manually edit some wm voxels, however with > no success. > Then I tried another solution by combining the FSL-FAST wm > segmentation with wm.mgz because I noticed that FSL_FAST > produced better results. Although the results did improve, > still the pial surface was incorrect. > The last option I tried was another set of parameters for > nu_correct. Because it concerns 3T scans I used > --proto-iters1000 and --distance 50 (see paper of Zheng et al, > NeuroImage 48, 73-83 (2009)) and previous e-mails on > this subject ("nu-correct parameters", 27 April 2010). > Also with these settings the pial surface chopped away > too many voxels. > I'm desperate, the manuals provide no options on how to fix > such problems. The pial surface can be corrected by only > REMOVING voxels that are included erroneously but there is > no way to extent the pial surface by ADDING voxels. > > Ed > > > > _______________________________________________ > Freesurfer mailing list > Freesurfer@nmr.mgh.harvard.edu > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer > > >