An examination of brain tumor detection using image segmentation methods

Abstract

Tumor segmentation from MRI data is a critical yet time-consuming manual task for medical experts, and our research aims to develop an effective algorithm to assist radiologists in diagnosing brain tumors. This study presents two parallel approaches for brain tumor detection: algorithmic and non-algorithmic. The research is divided into three phases: preprocessing and enhancement to remove film artifacts and noise using a weighted median filter, segmentation through block-based non-algorithmic methods and algorithmic methods such as Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO), and performance evaluation of these methods. Our analysis, conducted on 50 brain MRIs with expert-identified brain tissue classes, shows that the results closely match the radiologists' findings.