MRI脑部图像头骨剥离方法研究  

Skull-stripping MR Brain Images

作　　者：裴晓敏[1] 田秀华[1] 

机构地区：[1]辽宁工程技术大学电子与信息工程学院,辽宁葫芦岛125105

出　　处：《计算机仿真》2009年第3期205-207,共3页Computer Simulation

摘　　要：头骨剥离是脑组织分割过程中的关键环节。针对传统的MR I脑组织分割中头骨剥离速度较慢、分割精度不高的问题,提出一种基于水平集的头骨剥离方法。首先根据脑部图像灰度特征确定初始分割曲线,然后将曲线看成高维函数的零水平集,最后利用灰度和形态特征控制高维函数水平集演化,通过高维函数的演化确定曲线的演化,使曲线达到待分割图像边缘。对基于水平集的方法和现有的典型方法进行仿真,仿真结果表明该方法在分割精度上有明显提高。实验证明方法可精确地将脑部组织和非脑组织分离,达到头骨剥离的目的。Skull - stripping has become a key step in brain tissue segmentation. The traditional methods for MRI image skull - stripping are very slow and not precise. Aiming at this problem, the paper develops a method based on level set. Firstly, it defines the initial segmenting curve based on the gray characteristic of the image. Secondly, it judges the curve as the zero level set of the high - level function. Lastly, it evolves the high - level function by gray and morphological characteristics, then recognizes the edge of MRI brain images and segments the brain and non - brain image. Simulations of this method and the typical methods are carried out, the results show that this method can segment brain more precisely. The experiments show that this method can segment the brain and non - brain precisely, realize skull - stripping.

关 键 词：头骨剥离 脑组织 水平集 灰度 形态 非脑组织 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] R445.39[自动化与计算机技术—计算机科学与技术]