期刊名称：International Journal of Computer Science and Information Technologies
电子版ISSN：0975-9646
出版年度：2011
卷号：2
期号：6
页码：2573-2578
出版社：TechScience Publications
摘要：This chapter describes a method of segmenting MR images into different tissue classes, using a modified Gaussian Mixture Model. By knowing the prior spatial probability of each voxel being grey matter, white matter or cerebra-spinal fluid, it is possible to obtain a more robust classification. In addition, a step for correcting intensity non-uniformity is also included, which makes the method more applicable to images corrupted by smooth intensity variations. The segmentation is normally run on unprocessed brain images, where non-brain tissue is not first removed. This results in a small amount of non-brain tissue being classified as brain. However, by using morphological operations on the extracted GM and WM segments, it is possible to remove most of this extra tissue. The procedure begins by eroding the extracted WM image, so that any small specs of misclassified WM are removed. This is followed by conditionally dilating the eroded WM, such that dilation can only occur where GM and WM were present in the original extracted segments. Although some non-brain structures (such as part of the sagittal sinus) may remain after this processing, most non-brain tissue is removed.
