虚拟手术中基于质点弹簧建模及其切割技术研究

发布时间：2018-06-07 21:58

本文选题：虚拟手术 + 质点弹簧　；参考：《燕山大学》2015年硕士论文

【摘要】：随着科学技术的发展,虚拟现实技术得到了越来越多的实际应用。在医疗应用方面,通过采用计算机及其它设备的结合的虚拟手术系统进行虚拟手术仿真不仅可以减少手术训练成本,而且在降低手术风险、培养医疗人才等方面发挥着广泛的作用。本文从虚拟手术系统软组织建模与重建出发,以微创虚拟手术中人体软组织为研究对象,针对形状模型的建立、形变以及切割仿真等方面展开深入的研究,主要工作如下:首先,分析了现存各种几何及物理模型建模方法的优劣,比较其各自优缺点,为建模奠定基础。针对几何模型建立时图像分割效率较低,精度不高的问题,提出了利用Amira软件采用自动分割同手动分割相结合,自动分割继承手动分割细节信息的方式,得到真实度较高的人体软组织几何模型。针对质点弹簧模型数学基础,分析了形变的数学过程,建立符合程序设计要求的数据结构。其次,针对质点弹簧模型参数确定的随意性问题,提出了利用遗传算法确定质点弹簧模型参数的方法。采用遗传算法,利用模拟形变同参考形变间差异大小作为适应度函数求得模型参数的近似最优解,得到准确复现实际形变效果的质点弹簧形变模型。最后,对虚拟手术切割方面进行研究,提出四叉树与深度检测结合的切割起点搜索方法;利用模型单元重心间距表达的邻域信息建立索引搜索所有切割单元;以切割轨迹为基础,依据网格密度匹配进行伤口路径获取,针对获取的路径点采用区域向量加权完成伤口扩张实现整个切割操作过程,实现了较高仿真程度的伤口切割效果。
[Abstract]:With the development of science and technology, virtual reality technology has been more and more practical application. In the aspect of medical application, the virtual surgery simulation with the combination of computer and other equipment can not only reduce the cost of operation training, but also reduce the risk of operation. Training medical personnel and other aspects of play a wide range of roles. Starting from soft tissue modeling and reconstruction of virtual surgery system, this paper takes human soft tissue in minimally invasive virtual surgery as the research object, and carries out in-depth research on shape model establishment, deformation and cutting simulation, etc. The main work is as follows: firstly, the advantages and disadvantages of existing geometric and physical modeling methods are analyzed, and their advantages and disadvantages are compared to lay the foundation for modeling. Aiming at the problem of low efficiency and low precision of image segmentation when geometric model is established, this paper puts forward a method of combining automatic segmentation with manual segmentation by using Amira software, which inherits the detailed information of manual segmentation. The geometric model of human soft tissue is obtained. According to the mathematical basis of mass spring model, the mathematical process of deformation is analyzed, and the data structure that meets the requirements of program design is established. Secondly, aiming at the random problem of determining the parameters of mass spring model, a genetic algorithm is proposed to determine the parameters of mass spring model. By using genetic algorithm and using the difference between simulated deformation and reference deformation as fitness function, the approximate optimal solution of model parameters is obtained, and a mass spring deformation model which accurately reappears the effect of actual deformation is obtained. Finally, the research on virtual surgical cutting is carried out, and a cutting starting point search method combining quadtree and depth detection is put forward, and all cutting units are searched by using the neighborhood information expressed by the barycenter distance of the model unit. Based on the cutting path, the wound path is obtained according to the mesh density matching, and the region vector is used to complete the wound expansion to achieve the whole cutting operation process, and a high degree of simulation of wound cutting effect is realized.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：R616;TP391.9

【参考文献】