Abnormal Brain Tissue Isolation Using Image Segmentation Methods

Abstract

This paper describes a hybrid segmentation method for segmenting and isolating abnormal brain tissues or tumors in magnetic resonance imaging (MRI) images. A brain tumor refers to an abnormal proliferation of cells originating from brain tissue or surrounding structures. Nearby locations include the nerves, the pituitary gland, the pineal gland, and the membranes that cover the brain's surface. Brain tumor segmentation is a medical image analysis task involving separating brain tumors from normal brain tissue in MRI scans. Image segmentation methods are essential for accurately isolating abnormal brain tissue from MRI scans, enabling medical professionals and researchers to analyze and diagnose various neurological conditions. Different segmentation techniques were employed to segment and isolate the abnormal tissue in the brain. Thresholding methods were utilized to isolate the region of interest (ROI) using various threshold values (T). After experimenting with different values ​​for the threshold, some values ​​were chosen to be presented in this work as a sample of the method’s performance. The K-Means method yields varied segmentation outcomes depending on the parameter settings and the number of clusters (K). Only samples (K=6) are presented in this work. A hybrid segmentation method is proposed, combining thresholding and K-Means to produce a novel segmentation approach. The proposed method yields very good results, albeit with some unwanted regions. In the final step, an opening morphological operation was applied to refine the segmentation and enhance the clarity of the identified regions.