ON A NEW TECHNIQUE FOR GENERATING FINITE ELEMENT AND RAPID PROTOTYPED MODELS BASED ON 3D DATA

摘要

There has been increasing interest in the rapid and robust generation of three dimensional finite element meshes based on data obtained from a range of 3D medical imaging modalities. The authors have implemented 3D image segmentation techniques and have developed and implemented a new approach which generates finite element models of up to seven distinct regions of interest. The meshing is efficient, fast and robust - high quality mixed tetrahedral and hexahedral meshes can be generated for complex multiple structures in the matter of minutes and then exported to one of a number of commercial solvers. The proposed approach has a number of significant advantages when compared to previous techniques: (1) it is fully automated, (2) structures (segmented regions of interest) of arbitrary complexity can be meshed with pre-specified element qualities, (3) multiple structures can be meshed simultaneously and the mesh interfaces between these structures are guaranteed to be conforming - no gaps or overlaps (e.g. for knee joint), (4) where a relationship between signal strength and material properties exists, the material properties throughout a part can be based on the parent voxel's greyscale. A number of test cases will be presented as well as some experimental and numerical validation studies.