Issues with the Whole-Mesh Sampling Approach
While the described whole-mesh cortical depth sampling tool can be used to sample functional (or other) data sets at different relative cortical depth levels, the unregular layout of the vertices of a reconstructed mesh will lead to biased geometric sampling.
in the figure above, a small part of a high-resolution (> 1.25 million triangles) mesh has been extracted and zoomed in order to show a typical mesh-based cortical depth sampling; the zoomed views at 0.15 and 0.85 depth values on the right reveal that the spacing between vertices is irregular and that moving the same mesh to different cortical depth levels leads to systematic over- and undersampling of concave and convex regions. These issues, especially the irregular sampling, can be avoided by using the regular-grid cortical depth sampling approach.
Copyright © 2020 Rainer Goebel. All rights reserved.