Segmentation of White Matter Based on Region Growing and Threshold Theory

Abstract

In order to reduce the massive manual intervention involved in the existing segmentation methods for human cross-section slice images, a segmentation algorithm based on the theory of region growing and threshold in normal gray histogram was proposed in this paper, according to the features of slice images of human brain. More exactly, these slice images were initially segmented coarsely by means of the region growing. Then the method of threshold in normal gray histogram was adopted to refine the segmentation. The experimental results indicate that the proposed algorithm can segment white matter accurately and effectively.