基于医学图像序列生成四面体网格

Tetrahedral Mesh Generation from Medical Sectional Images

为实现医学图像的几何建模,提出了从一个医学图像序列生成三维四面体网格模型的方法。该方法按分层思想实现四面体网格的生成:首先在各医学图像上提取待重建器官组织的平面轮廓,并将其离散化;然后对各平面轮廓区域生成二维三角网格;最后分别在各相邻断层之间采用三遍扫描法连接四面体单元,从而得到三维四面体网格模型。用该方法重建人体大腿膝部的几何模型,得到一个含1468个节点与4378个四面体单元的体网格,该网格被应用于虚拟手术系统的切割形变等有限元分析与计算。

A method of tetrahedral mesh generating from a series of medical images is proposed to reconstruct anatomic geometrical models. Tetrahedral meshes are generated in a layered way as follows： firstly,planar contours for the human organs or tissues to be reconstructed in each medical image are extracted and then discretized,secondly,2D triangular meshes are generated according to these discretized contours on respective image layer,and finally,a set of tetrahedral elements are linked between every two adjacent layers by three-scanning means,which constitute the final 3D mesh of several layers. Example of geometrical modeling by using the presented method is presented,which constructs a tetrahedral mesh of the human leg near the genual part with 1468 nodes and 4378 elements. The generated mesh is applied effectively to cutting deforming simulation in virtual surgery.