Affiliations: School of Electrical, Electronic and Mechanical Engineering, University College Dublin, Dublin, Ireland | School of Human Kinetics, University of Ottawa, Ottawa, ON, Canada
Note: [] Currently at Ginkgobioworks.com, Boston, MA, USA.
Note: [] Address for correspondence: Michael D. Gilchrist, School of Electrical, Electronic and Mechanical Engineering, University College Dublin, Dublin 4, Ireland. E-mail: [email protected].
Abstract: This paper describes the development of a new automatic mesh generator for complex biological objects based on voxel hexahedral meshing (VHM). The system produces hexahedral meshes with multiple material classifications from three-dimensional (3D) computed tomography (CT) datasets. The quality of the VHM meshes is improved using a series of mesh connectivity improvement algorithms, leading to the application of a new and novel boundary smoothing algorithm. The boundary smoothing algorithm classifies boundary nodes into different groups based on the configuration of their neighbouring elements: this provides a varying compromise between smoothness and element distortion. The high anatomic detail of which the method is capable is shown for meshes of single and mixed material types. This system illustrates how to harness the speed and simplicity of VHM as an initial mesh generator while simultaneously improving the quality of those meshes with additional meshing techniques. MATLAB code for this software is freely available for research purposes upon request.
Keywords: Mesh generation, Laplacian smoothing, voxel mesh, hexahedral element
