Hi Martin,

can you show the orig surface also? This is probably due to a topological defect being incorrectly fixed. You can load the ?h.orig.nofix (or inflated.nofix) surface and read in the ?h.defect_labels curvature file to see if there was a defect in these regions. Does the white matter connect across the banks of the sulcus if you go forward and back a few slices? Or is it disconnected? The wm looks extremely bright there, not sure why. Did you pick any control points?

Bruce

On Tue, 11 Apr 2006, Martin Ystad wrote:

Hi, I'm running the stable release v 3.0.1 for centos4 and I've had some problems with the segmentation. The main surface doesn't follow the wm-surface in some of the sulci, it rather 'jumps' across to the next gyri. (see picture) I've tried to edit the wm-surface and running autorecon2 again, but it doesn't work. Any ideas?

Thanks,

Martin Ystad University of Bergen, Norway.