CASIA OpenIR  > 毕业生  > 硕士学位论文
Alternative TitleMagnetic Resonance Imaging of Brain Segmentation based on K-means Clustering Graph Cuts
Thesis Advisor田捷
Degree Grantor中国科学院研究生院
Place of Conferral中国科学院自动化研究所
Degree Discipline计算机应用技术
Keyword图割 脑部mri 高斯混合模型 K均值聚类 Graph Cuts Magnetic Resonance Imaging Of Brain Gmm Kmc
Abstract图割(graph cuts),是2001年由Boykov和 Jolly提出一种同时基于区域和边界的交互式图像分割算法。目前,已成为在计算机视觉的很多领域都非常流行的一种全局最优化算法。 很多疾病的发生与病情变化,都与脑部各种组织结构的体积改变有关。这就需要我们把脑部核磁共振图像中各个组织,如灰质、白质和脑脊液精确地进行分割。本文中,主要贡献如下: ① 针对传统图割方法分割脑部MRI时的不理想结果,提出了基于K均值聚类(KMC)的graph cuts方法,通过用在脑部MRI的仿体和真实数据上进行实验,证明了该方法的有效性和高效性; ② 详细阐述了分割脑部MRI的经典方法,即基于模糊C均值,和基于马尔科夫随机场的方法,并且指出这两种方法在分割MRI时的不足; ③ 把传统的graph cuts分割方法和改进的方法,集成到MITK框架中,并且也方便把其它模块集成到graph cuts方法中。 全文共分六章。第一章,介绍了MRI的发现和使用,以及MRI的成像原理,以及研究脑部MRI的分割算法的意义;第二章,系统介绍了分割脑部MRI的经典算法,并指出它们的不足;第三章,介绍了graph cuts的发展历程,并详细阐述了graph cuts图像分割算法;第四章,介绍了基于KMC的图像分割算法,并在仿体和真实数据上进行了实验;第五章,介绍了对医学图像分割进行评价的意义,以及经典评价指标;第六章,对本文的工作进行回顾,并对脑部MRI分割方法以及图像分割评价方法进行了展望。
Other AbstractGraph cuts, is a segmentation method based on boundary and region, and was proposed by boykov and jolly in 2001. It has been a globally optimal method in many aspects of computer vision. Progress or remission of various diseases is related to the volumetric analysis of different parts of the brain. It needs the segmentation of brain magnetic resonance images to the main tissue types: white matter(WM), gray matter(GM) and cerebro-spinal(CSF). In this paper, the main contributes includes following tissues: ① Proposed graph cuts segmentation methods based on K- means Clustering, to resolve the poor results using the conventional graph cuts.Its evaluation was performed using both phantoms and real Magnetic Resonance Imaging of brain, showing the effectiveness of the proposed method. ② Described the classic segmentation methods of Magnetic Resonance Imaging of brain, and pointed the drawbacks of the two methods; ③ Integrated the conventional graph cuts and proposed methods into Medical Imaging Algorithm Tookit(MITK),and it is convenient to integrate other modules to the algorithm. The thesis is divided into six chapters. Chapter one introduces the discovery, use and fundamental principles of MRI. In chapter 2, we described the classic segmentation methods of Magnetic Resonance Imaging of brain, and pointed the drawbacks of the two methods. We introduce the graph cuts segmentation method in detail in chapter 3. In chapter 4, we propose graph cuts based on KMC, and demonstrate the effective of the method. We introduce the significance of evaluation of segmentation methods, and classic evaluation criteria. We review the work of this thesis in chapter 6.
Other Identifier200728017029234
Document Type学位论文
Recommended Citation
GB/T 7714
吴永芳. 基于K均值聚类的图割脑部MRI分割算法研究[D]. 中国科学院自动化研究所. 中国科学院研究生院,2010.
Files in This Item:
File Name/Size DocType Version Access License
CASIA_20072801702923(799KB) 暂不开放CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[吴永芳]'s Articles
Baidu academic
Similar articles in Baidu academic
[吴永芳]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[吴永芳]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.