医学仿真中一种高效的生物组织几何建模方法

Effective Tissue Geometry Modeling Method in Biomedical Simulation

在生物医学仿真系统的应用背景下,以医学体数据为数据来源,提出一种高效的生物组织几何建模方法。该方法包括等值面提取、等值面预处理、Delaunay表面重建和四面体填充四个部分。引入基于二义性检测索引表的Marching cubes算法,保证生成等值面拓扑的正确性;提出三角形顶点射线法去除等值面简化后内部遗留的冗余网格;引入顶点控制球策略,克服限定Delaunay细分算法无法收敛的缺陷,有效生成Delaunay表面模型;提出自适应的内部节点添加技术进行Delaunay四面体填充。实验表明,提出的方法在有效保证网格拓扑的基础上,可根据需求生成满足Delaunay准则的目标组织几何模型。生成的模型已应用于生物医学仿真系统。

A reliable and effective modeling method was proposed for Biomedical Simulation. The method takes medical volume data as input and consists of four parts - iso-surface generation, preprocess, Delaunay surface re-meshing and tetrahedralization. An improved marching cubes method based on ambiguity detection index table was presented to generate tissue iso-surface. Redundant triangles from the simplified iso-surface were deleted by clever strategy, then a polyhedron surface was produced. To construct Delaunay surface, a sphere control based Delaunay refinement was introduced to re-meshing the polyhedron. For the demands of tetrahedral mesh, space-disassembling method combining with random perturbations was presented to fill Delaunay tetrahedrons. Experimental results show that the presented method can produce tissue geometry model conforming to the Delaunay criterion. The generated meshes are well suited for Biomedical Simulation.