A method for segmentation of CT head images

Abstract

A novel method for automatic segmentation of computed tomography (CT) head images of patients having spontaneous intracerebral brain hemorrhage has been presented in this work. The method consists of four major phases. The first phase performs a brightness normalization of a CT image by applying a K-means clustering algorithm to the pixel brightness values. A feature extraction based on a special receptive field is done in the second phase. The third phase performs a pixel classification by means of a feed-forward error-back propagation neural network. In the last phase, a rule-based expert system is used to perform image labeling. The proposed method has been applied to real patient CT images and has shown encouraging results.