基于自动策略的脑图像分割  

A brain segmentation algorithm based on automatic strategy

作　　者：曹会志[1] 罗述谦[1] 

机构地区：[1]首都医科大学生物医学工程学院,北京100069

出　　处：《中国医学影像技术》2007年第2期303-306,共4页Chinese Journal of Medical Imaging Technology

基　　金：国家自然科学基金资助项目(60472020)。

摘　　要：目的脑图像分割在外科手术规划和脑疾病诊断等方面都起着极为重要的作用,建立脑图像分割的自动策略成为一种需要。方法通过各向异性滤波,统计阈值分割,数学形态学滤波,和基于模糊连接算法对脑图像进行自动分割。结果实验表明这种分割策略能取得良好的分割结果。结论本文提出的算法可以有效地完成脑图像的自动分割工作。Objective Brain image segmentation is of utmost importance in planning surgical procedures and diagnoses of brain diseases, and the segmentation based on automatic algorithm is needed. Methods This method integrate anisotropic diffusion filter, statistical threshold, mathematical morphological operations, and algorithm of fuzzy connected is used in the process of segmentation. Results Datasets have been used to test our algorithm, and results show that our method is effective. Conclusion The algorithm used in this paper is very useful in automatic segmentation of brain image.

关 键 词：医学图像 图像分割 模糊连接 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]