Description
December 17, 2011



The following MATLAB code will load the both outer and inner cortical surface mesh format into MatLab. It is tested on MATLAB 2009 (version 7.9, 64-bit Mac intel). The complete data set is used in studies [1] [2] [4]. If you want to use the data set for publication, please contact Moo K. Chung.  

load autism.mesh.coord.mat

The data set contains the both inner and outer cortical surfaces of 16 autistic and 11 control subjects. In the mat file, the variable autism_inner is 3 dimensional array of size 16 x 3 x 40962. Each dimension indexces subject number, (x,y,z) coordinates and coordinate values. To visualize the inner surface of the 1st autistic subject, need to reformat the data structure.

surf.vertices=squeeze(autism_inner(1,:,:))';
surf.faces=tri;
figure; figure_wire(surf, 'yellow', 'white');

The outer surface is visualized as follows:

surf.vertices=squeeze(autism_outer(1,:,:))';
surf.faces=tri;
figure; figure_wire(surf, 'yellow', 'white');

To visualize the y-coordinate values on the outer surface

figure; figure_trimesh(surf,surf.vertices(:,2))

The cortical thickness for the first autism subject is computed as

thickness=squeeze(L2norm(autism_outer(1,:,:) - autism_inner(1,:,:)));
figure; figure_trimesh(surf,thickness)

The cortical thickness values are highly noise so we need to perform heat kernel smoothing. See [1] and [2] for the detailed implemenation of heat kernel smoothing. We can also extract cortical thickness values at vertex 10 for all subjects

thickness1=L2norm(autism_outer(:,:,10) - autism_inner(:,:,10))
thickness2=L2norm(control_outer(:,:,10) - control_inner(:,:,10))
figure;
plot(thickness1,'or');
hold on;
plot(thickness2,'ob');

MNI surface mesh format
May 9 2006.

The following MATLAB code will load MNI cortical mesh format into MatLab. It only read ASCII-based MNI format. MNI surface mesh format  always ends with *obj extension. This is based on MATLAB 6.1. implementation. The MNI cortical mesh datat set has been used in the studies [1-4].

• mni_getmesh.m loading triangle MNI mesh format into MATLAB.
• Outsurface.obj sample cortical surface mesh.
• sphere.obj cortical surface flattened onto a sphere.

[tri,coord,nbr,normal]=mni_getmesh('outersurface.obj');

load thickness.data;

output=hk_smooth(thickness',tri,coord,nbr,1,200);

trisurf(tri, coord(1,:),coord(2,:),coord(3,:),output);

• coord (3x40962) gives Cartesian coordinates of 40962 nodes.
• nbr (40962x6) gives neighboring node indices. Each node has up to 6 neighboring nodes. First 12 nodes have only 5 neighbors while the remaining 40962-5 nodes have 6 neighbors.
  
• normal (3x40962) gives the normal vectors of the cortex.
• tri (81920x3) gives node indices that forms each of 81920 triangles.

References
December 17, 2011

1. Chung, M.K., Robbins, S., Evans, A.C. 2005. Unified statistical approach to cortical thickness analysis. Information Processing in Medical Imaging (IPMI). Lecture Notes in Computer Science (LNCS) 3565:627-638. Springer-Verlag.
2. Chung, M.K., Robbins,S., Dalton, K.M., Davidson, Alexander, A.L., R.J., Evans, A.C. 2005. Cortical thickness analysis in autism via heat kernel smoothing. NeuroImage 25:1256-1265.
3. Shen, L., Chung, M.K. 2006. Large-scale modeling of parametric surfaces using spherical harmonics. Third International Symposium on 3D Data Processing, Visualization and Transmission (3DPVT).
4. Chung, M.K., Bubenik, P., Kim, P.T. 2009. Persistence diagrams of cortical surface data. Information Processing in Medica Imaging (IPMI). Lecture Notes in Computer Science (LNCS). 5636:386-397.