Hi Kai, you can use mris_expand to generate a "halfway" surface. The thickness is the average of the distance to the nearest point from the white to the pial and from the pial to the white, so it's somewhat different from the mid surface.

cheers, Bruce

On Mon, 8 Jun 2009, Kai Lutz wrote:

Dear List,

in an attempt to divide grey matter segments into an outer and inner layer, I am currently using the vertex coordinates for pial surface (in Matlab>> read_surf('lh.pial') ) and vertex coordinates for white matter surface (MATLAB>>read_surf('lh.white') ). Perhaps naively, I assume, that the order of the two coordinate matrices is identical and calculate the wanted coordinates of a point on a line (say half way) in between the corresponding pial and white matter vertices.

What puzzles me is that the distances between corresponding (?) white matter and pial surface vertices are not equal to the cortical thickness values given by >>read_curv('lh.thickness') (GM-WM distances up to 9.3 mm are found, whereas cortical thickness values vary between -0.24 and 5 mm in the data set I am working on)

--> To determine a layer, separating outer from inner half of the cortex, would it be a more valid to A: determine vertex normals of pial surface and travel along them by a distance 1/2 the cortical thickness? (Q: which thickness value belongs to which vertex? where do the negative thickness values come from?) or B: determine points half way in between vertex coordinates of corresponding pial and WM vertices? (Q: how can I assure their correspondence?)

Is anyone working on / has solved a related problem? Any hint would be greatly appreciated!

Cheers, Kai