the intensity normalization has a region growing component that is bounded by intensity gradients. In images like MPRAGE, where SNR is traded for CNR it can help to raise the threshold for stopping so that more white matter is included (that's what the -b 20 does, I think the default it 10 or 15). This is usually not good for SPGR/FLASH and can result in the region growing "escaping" the white matter and eroding the gray matter, which looks like it happened in your case.
cheers Bruce
On Tue, 13 Nov 2012, youngmin huh wrote:
Thanks. I'll try without expert opts.But i'm not sure if i can get better results. Before I found out these expert opts, I ran some images recon2 without expert opts. And they got poor segmentation. Actually, some images showed improvement with those options. I'm curious why these options work for some images and don't for others. And, 3D SPGR sequences were used. (TR=22, TE=4, thickness=1.4, FOV=240*240, FA=40)
2012/11/12 Bruce Fischl fischl@nmr.mgh.harvard.edu Can you try it without the expert opts? And what sequence is the data?
On Nov 12, 2012, at 3:35 AM, youngmin huh <ymin1123@gmail.com> wrote: > Hi > I'm working with some images, and freesurfer didn't properly detect some of their white matter boundaries. > I attached some examples. > I ran recon2 with expert options(-b 20 -n 5). > As you can see, they have inaccurate white matter boundaries in curved area. > I think it's not that highly curved, though. > Any comment will be appreciated. > > Thanks, > > Youngmin Huh ><image.jpeg><image.jpeg> _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.