Ischemic Stroke Identification by Using Watershed Segmentation and Textural and Statistical Features

Abstract

The algorithm presented in this paper identifies the ischemic stroke from CT brain images by extracting the textural and statistical features. Our algorithm starts by preprocessing of our CT images, and then image enhancement is performed. The brain CT images are segmented by Marker Controlled watershed. We obtained the Grey Level Co-occurrence matrix (GLCM) to extract the textural and statistical features. The experimental results showed that the over-segmentation due to noise is resolved by Marker controlled watershed. The textural and statistical features showed that the values of contrast, correlation, standard deviation and variance of normal CT images are less than those of abnormal CT images (contains ischemic stroke), where the values of homogeneity, energy and mean are bigger in normal CT images than those of abnormal CT images.