Hi Shani
we described in in the recon1 NeuroImage paper (1999) and the PNAS paper, I think the next year. We don't impose any constrains on the pial surface as it evolves from the white surface (e.g. "link" terms in the energy functional), but let the vertices move whereever they want to in order to find an image intensity that is what we estimate the gray/csf boundary to be in that region.
Can you tell us a bit about what you want to do with the normals? We typically don't use them for sampling of hires data for example, but rather use mris_expand to generate a series of intermediate surfaces.
cheers Bruce
On Sun, 23 Oct 2011, Shani Ben Amitay wrote:
Dear freesurfers, In order to generate cortical vector normals connecting homologous vertices of the inner and outer surface, I would like to better understand how the gray matter/csf surface is reconstructed and how correspondence between inner and outer surfaces is achieved.
Does anyone can explain or give a reference?
Thanks , Shani