Hi Gregory,

it's hard to tell accuracy looking at one slice, but that one looks fine. Note that we do not generally use the cortical segmentation in the aseg - for that we use the surfaces. So if you want to assess accuracy you should overlay the ?h.white and ?h.pial surfaces and see if cortical gray matter is accurately contained between them.

cheers, Bruce On Thu, 9 Jul 2009 simon@cyceron.fr wrote:

Hi.

I'm a new user of freesurfer. The default settings of recon-all performed a good white segmentation of my data but I find that grey matter (see attached picture) could be optimized when I look at aseg.mgz results. Indeed the surface cortical grey matter was not extend as expected. What were the different parameters that I can optimize to have a better grey matter segmentation without manual correction?

Thank a lot for your answers.

Gregory SIMON