【作者】 王媛媛；

【导师】 宋小文； 黄长林；

【作者基本信息】 浙江大学 ， 机械设计及理论， 2010， 硕士

【摘要】 近年来,随着计算机断层扫描(CT:Computed Tomography),核磁共振成像(MRI:Magnetic Resonance Imaging),超声(US:Ultrasonography)等医学成像技术的产生并且应用越来越广泛,而医疗仪器获得的二维图像却远远满足不了要对数据进行精确分析的要求。因此,医学图像的三维重建及可视化有重要的研究意义和应用价值。本文结合具体医学图像,以自适应细分为网格简化方法,VTK可视化工具包为可视化实现方式,主要研究内容如下：介绍三维重建技术的国内外发展状况,整理和总结三维重建的基本理论,阐述三维数据的建模方法,分析医学图像的预处理、图像的分割及模型表面网格简化。讨论CT、MRI图像的获取并实现数据的输入,针对数据探讨图像的滤波及断层图像间插值与断层间表面重构。介绍体素模型与等值面的定义,并对等值面的抽取方法移动立方体法(Marching Cubes)算法和移动四面体(Marching Tetrahedra)算法的基本原理进行详细介绍和对比。由于采用MC和MT算法抽取等值面构建的表面几何模型所包含的三角面片数量较大,难以实现实时绘制显示。本文引入自适应细分方法进行网格模型的网格简化,在保持原模型的特征和较好的视觉效果的前提下,简化模型网格,大大的提高了实时交互绘制能力。讨论基于可视化分析的研究模型,并分析可视化的流程。借助于VTK可视化工具包,采用基于面绘制的Marching Cubes算法,处理DICOM标准的二维医学图像数据,并进行三维重建,真实的再现了物体的三维面貌,同时可以交互的进行旋转、平移等操作。国内常用的三维可视化工具类似OpenGL,存在代码执行率低或计算能力差等缺陷,VTK避免了这类问题。论文最后综述了医学图像三维重建系统的开发,分析了系统需求,介绍了算法和主要流程,提出并建立了系统数据结构。在VC++6.0开发环境下,使用VTK工具包和MFC库开发出一个医学图像三维重建及可视化系统,可以读取特定格式的医学图像数据,进行三维重建,及最终的实时交互。经测试本文以上算法正确,且重建效果良好,调整方便,可用于各种DICOM标准医学图像数据的三维重建及可视化。更多还原

【Abstract】 Recently, as the medical imaging technology like Computed Tomography、Magnetic Resonance Imaging、and Ultrasonography has came into being and promote a promising future. The medical equipment acquired two-dimensional image can hardly satisfy the data requirements for accurate analysis, so the three-dimensional reconstruction of medical images has an important significance on science and worthiness in practical application.Describes the development of three-dimensional reconstruction technology at home and abroad, collects and sums up the three-dimensional reconstruction’s basic theory, expounds the three-dimensional data modeling methodology, analysis the medical image’s preprocessing, image segmentation and model surface mesh simplification.Discuss the capture of CT, MRI image and data input. For the data to explore the image filtering and inter-sectional image interpolation and surface reconstruction between inter-sectional. Introduction voxel model and the definition of iso-surface. Contrast the extraction method of iso-surface for Marching Cubes algorithms and Marching Tetrahedra algorithm. The basic principles also been descripted full of detail.3D surface model reconstructed with MC or MT algorithm contains huge number of triangles. It is difficult to render them in real time. This paper introduces the method of adaptive mesh subdivision to simplify the original mesh data. Simplify the model grid while maintaining the characteristics and visual effects of the original model. Greatly improve the real-time interactive rendering capabilities.The discussion of the research model is based on visual analysis. Also analyzes the visualization process. With VTK toolkit and based on surface rendering of the Marching Cubes algorithm, deal with the DICOM standard two-dimensional medical image data, then reconstruct a three-dimensional image of the data. Representation the real three-dimensional image of the object. Can interact such as rotate, translation and other operations at the same time. OpenGL is commonly used as three-dimensional visualization tools at home, however, the coding execution rate is low or poor numeracy, VTK avoids such problems.This paper finally summarizes the development of three-dimensional medical image reconstruction system, analysis the requirements of system, describes the algorithm and the main processes, designs and establishes a system data structure. Under the environment of VC++ 6.0, the VTK toolkit and MFC libraries is used to develop a three-dimensional medical image reconstruction and visualization system, which can read the medical image data, reconstructe three-dimensional imge and realize real-time interaction. Through plentiful practical example,the system in this thesis is proved to be right and feasible.更多还原