Brain Tumor Detection of Skull Stripped MR Images Utilizing Clustering and Region Growing

Abstract

Brain tissues segmentation is usually concerned with the delineation of three types of brain matters Grey Matter (GM), White Matter (WM) and Cerebrospinal Fluid (CSF). Because most brain structures are anatomically defined by boundaries of these tissue classes, accurate segmentation of brain tissues into one of these categories is an important step in quantitative morphological study of the brain. As well as the abnormalities regions like tumors are needed to be delineated. The extra-cortical voxels in MR brain images are often removed in order to facilitate accurate analysis of cortical structures. Brain extraction is necessary to avoid the misclassifications of surrounding tissues, skull and scalp as WM, GM or tumor when implementing segmentation algorithms. In this work, two techniques have been implemented to extract the brain tissues as elementary step. The next step was utilizing the resultant skull stripped images as input of four segmentation algorithms to extract the tumor region and calculate the area value of it. The resultant skull stripped images for complete set of T2-weighted images and the adaptive K-Means clustering techniques proved the robust performance of these proposed algorithms.