基于形变模型的医学影像分割的研究与应用
文献类型：学位论文
| 作者 | 朱付平 |
| 学位类别 | 工学博士 |
| 答辩日期 | 2003-05-01 |
| 授予单位 | 中国科学院研究生院 |
| 授予地点 | 中国科学院自动化研究所 |
| 导师 | 田捷 |
| 关键词 | 医学影像分割 参数形变模型 几何形变模型 Level Set方法 灌注成像 医学影像处理与分析 分水岭方法 Fast Marching方法 Medical Image Segmentation Parametric Deformable Model Geometric Deformable Model Level Set Method Perfusion Imaging Medical Ima |
| 其他题名 | Study and Application of Medical Image Segmentation Based on Deformable Model |
| 学位专业 | 模式识别与智能系统 |
| 中文摘要 | 医学影像学给传统的医学诊断和治疗方式带来了翻天覆地的变化，特别是 随着计算机图形学、模式识别、人工智能、虚拟现实和计算机网络等技术在医 学影像领域的应用，逐渐形成了一门交叉学科——医学影像处理与分析。医学 影像分割是医学影像处理与分析中的关键技术和难点，分割后的影像正被广泛 应用于各种场合，如组织容积的定量分析，辅助诊断，病变组织的定位，解剖 结构研究，治疗规划，功能成像数据的局部体效应校正和图形引导手术。 本文的研究重点是医学影像的分割方法。由于成像设备和手段的不同，与 其他影像相比，医学影像具有形状复杂多样、个体问差异大、信号不均匀、边 缘模糊、多噪声等特点，因此医学影像的自动分割十分困难，至今也没有一种 通用的分割方法能满足不同应用的需要。另一方面，由于医学影像的分割往往 需要处理多张切片，由医生手工逐一勾勒出每张切片中的对象边缘不仅十分繁 琐，而且准确性完全依赖于医生的影像专业经验。因此有必要在分割中引入专 家的作用，将专家的辨识能力和计算机的描绘能力相结合。 本文的研究正是在人机交互的思想指导下，针对医学影像的特点和从临床 应用的需求出发，提出了几种交互式的医学影像分割方法，并在实验室自主开 发的三维医学影像处理与分析系统上加以实现和验证。本文的主要工作包括： （1）提出了一种基于参数形变模型的医学影像分割算法：首先通过Waterslled 算法对医学影像进行过度分割，获得每个小区域的边缘后，改进的Live Wire算 法可以获得单张切片的准确分割，将其映射到下一张切片，作为主动轮廓模型 的初始轮廓，在狄度模型的作用下，自动收敛到目标的边缘。我们通过引入轮 廓的活动区域的概念，限制了轮廓点的演变范围。实验表明，我们的算法缩小 了能量最小化过程的复杂性，减少了用户干预的次数，提高了分割的准确度。 （2）提出了一种改进的Fast Marching方法，旨在克服传统的Fast Marching 方法在处理模糊医学影像边界方面的不足。其主要思想是：首先引入过度分割， 得到医学影像的过度分割图；接着通过计算过度分割图中相邻区域之间的相似 度，重新定义了Fast Marching方法的速度函数；此外，我们还引入了堆数据结 构来保存像素点到用户指定的种子点的到达时间，引入了区域邻接图数据结构 来保存过度分割后的区域。实验表明改进后的算法一方面可以很好的收敛到模糊的医学影像边缘，另一方面也加快了Fast Marching方法的遍 |
| 英文摘要 | From 1970's, CT, MR, PET and other medical imaging instruments had been successfully applied to clinic medicine. With the development of computer graphics, pattern recognition, artificial intelligence, virtual reality and computer networking, a new branch of research medical imaging process and analysis is coming into being and in the ascendant. The emphasis of this dissertation is the research on medical image segmentation, which is one of the bottlenecks of medical image processing and is a fundamental building block for higher-level image analysis. Image segmentation remains a difficult task, however, due to both the tremendous variability of object shapes and the variation in image quality. In particular, medical images are often corrupted by noise and sampling artifacts, which can cause considerable difficulties when applying classical segmentation techniques. As a result, none of the method that had good result for general images had been proposed up to now. In practice, in order to obtain the boundary of a 3-D object, there's usually tens, even hundreds of slices to be segmented. It is very tedious that an expert anatomist manually delineates the boundaries of different structures. Then it is necessary to introduce the interaction of experts to the segmentation algorithm. Under the guidance of this idea and aiming at the character of medical imaging, several interactive segmentation methods are proposed for specific medical applications in this dissertation. The main work of this dissertation is as follows: (1) A medical image segmentation algorithm is proposed based on parametric deformable model. First, the traditional live wire algorithm is modified by combining it with the watershed method, and one slice in a medical slice series is segmented accurately by live wire algorithm. Then, the segmented results are mapped to next slice as the initial contours of active contour model. Last, the computer will segment the nearby slice using the modified active contour model based on gray model. We introduce the concept of active contour region to confine the propagating scope of contour points. The experiment results show that our algorithm can reduce the complexity of the minimization process and recover the boundary of the desired object reliably with only little user intervention. (2) A modified fast marching methed-based approach is proposed for correcting the traditional fast marching method'S deficiency in processing the fuzzy medical image boundary.First,the over-segmented map is acquired by over-segmentation method.Next,the statistical similarity measure of the nearby regions is calculated and the traditional fast marching method is modified by redefining the speed function based on statistical similarity measure.Moreover,by introducing heap data structure to save the arrival time of pixel to seeded point,and region adjacency graph(RA G)to describe the over-segmented regions.The experim |
| 语种 | 中文 |
| 其他标识符 | 748 |
| 源URL | [http://ir.ia.ac.cn/handle/173211/5754]  |
| 专题 | 毕业生_博士学位论文 |
| 推荐引用方式 GB/T 7714 | 朱付平. 基于形变模型的医学影像分割的研究与应用[D]. 中国科学院自动化研究所. 中国科学院研究生院. 2003. |
入库方式： OAI收割
来源：自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。