Multi-Tissue Tetrahedral Mesh Generation from Medical Images

A method extended from a constrained Delaunay algorithm, which generates the multi-tissue tetrahedral meshes for segmented medical images, is proposed in this paper. With an external boundary surface and internal sampling points, the anatomical objects are tetrahedralized homogenously. A new tissue-labeling scheme is utilized to assign the distinct tissue types of the additional inserted points and a division scheme is introduced to partition mixed-tissue tetrahedrons. The multi-tissue mandible tetrahedral models based on the Chinese Visible Human (CVH) dataset have been built with the proposed method. The results have the uniform shapes and good property for interpolation. Furthermore, the boundaries between different tissues are also easily extracted.