节点文献

基于图像的MCNP数字人体建模与仿真研究

Study on Image-based MCNP Modeling and Simulation of Digital Human

【作者】 施灿辉

【导师】 刘晓平;

【作者基本信息】 合肥工业大学 , 计算机软件与理论, 2003, 硕士

【摘要】 随着人体解剖学和肿瘤放射物理学的逐渐发展,以医学序列图像(如CT、MRI)为基础进行的计算机辅助诊断治疗正发挥着越来越大的作用,放疗计划制定的优劣取决于很多因素,现有放疗技术中最先进的是适形调强放疗技术,但依然存在CT模拟定位、体位固定以及剂量计算不精确等问题。目前国内商用放疗计划软件在模拟计算剂量分布的时候,大多采用水体模加修正因子的方法,如果使用大型仿真程序MCNP进行精确剂量计算,需要建立精确的人体体模,同时会带来大量的时间消耗。本文在初步研究使用VTK(可视化软件包)进行医学序列图像可视化之后,集中研究了基于图像的人体建模方法以达到使用MCNP进行精确剂量计算的目的。VTK是一种用于3D计算机图形学、图像处理及可视化的工具。医学序列图像的可视化技术包括体绘制和表面绘制两大类,各有其特点和适用领域。使用VTK可以很大程度地简化医学序列图像重建的问题。MCNP是大型蒙特卡罗中子、光子和电子输运仿真程序,是国际上用来进行剂量计算的重要工具之一。本文主要工作有以下几点:(1) 初步研究了医学序列图像可视化技术的方法、原理、特点以及VTK实现三维重建及可视化的算法过程。(2) 研究并实现了一种从医学序列图像CT建模并使用MCNP进行精确剂量计算的方法,用于建立基于CT值的精确人体脑部模型。文中详细介绍了其中核心算法—“体元合并”的两种实现方法。(3) 该算法生成的MCNP输入文件用以进行剂量计算,解决实际应用中计算时间过长和MCNP程序栅元数量限制的问题,实验结果证明了算法的正确性、有效性。

【Abstract】 As the development of human anatomy and radiation oncology physics, computer aided diagnostic therapy which is based on medical images (such as CT, MRI ) is more and more useful in practice. There are several ways of radiation therapy, and the most advanced way is the technique of intensity modulation radiation therapy. The way exists some problems such as orientation of CT machine, fixation of human body, and imprecision of dose calculation, etc. The patient model adopted in the existing commercial radiotherapy planning system is based on water material with inhomogeneities and skin curvature correction. Accurate patient model is needed for dose calculation using MCNP, and the calculation will be time-consuming. The author elementarily studied how to finish visualization of medical images by using VTK, and primarily studied how to build the patient model to make MCNP calculate more precisely.The Visualization ToolKit (VTK) is an open source, freely available software system for 3D computer graphics, image processing, and visualization used by thousands of researchers and developers around the world. The visualization technique of medical images includes two parts: volume rendering and surface rendering. They have respective characteristics and application fields. It could greatly simplify the problem about reconstruction of medical images to use VTK. MCNP is a general-purpose Monte Carlo N-Particle code that can be used for neutron, photon, electron, or coupled neutron/photon/electron trans-port, including the capability to calculate eigenvalues for critical systems.The main contribution of this dissertation includes:1. Study of two rendering methods and the algorithm of visualization of medical images by using VTK.2. Introduction of an image-based brain model which can be applied in dose calculation using MCNP. The key algorithm in MCNP modeling named as "voxel uniting" contains two methods.3. The algorithm solves the problem of time consuming and limit of the amount of cells, and its function & validity are proved by some examples.

【关键词】 三维重建体元栅元DICOM蒙特卡罗
【Key words】 3D reconstructionVoxelCellDICOMMCNP
  • 【分类号】TP391.9
  • 【被引频次】7
  • 【下载频次】618
节点文献中: