do any of the label entries have -1 for a vertex number in the first column? If so, tksurfer will sample it onto the surface finding the nearest surface vertex.
On Sun, 3 Jun 2012, David Groppe wrote:
Hi FreeSurfers, I am attempting to plot a label in MATLAB using the function read_label.m. The label was produced in tksurfer (see label_tksurfer.jpg) from a mask created by FSL. When I try to import the label into MATLAB via read_label.m, I get the correct number of nonzero vertices (i.e., the same number reported by tksurfer when I load the label). However the resulting plot is spotty (see label_matlab.jpg). I suspect that some vertices that are colored as the label in tksurfer are not actually part of the label file (i.e., tksurfer is automatically doing some filling in that I'm not aware of). Is this true? Any tips on how to solve the problem would be greatly appreciated. much thanks, -David