Latest Trends in Automatic Glioma Tumor Segmentation and an Improved Convolutional Neural Network based Solution

Abstract

A Brain tumor is an abnormal cell growth in the brain tissues, these tumors are difficult to treat and severely affect the patient's cognitive ability. Out of all brain tumors, gliomas are the deadliest with the least survival rate. The focus of brain tumor segmentation task is to separate tumor tissue such as edema, tumor core from the healthy tissues i.e. white cells, Cerebrospinal Fluid and gray matter. Manual diagnosis of brain tumors from a large amount of patient's MRI images is a tough and time-taking process. With the advent of new approaches, automatic segmentation processes are becoming more effective and clinically accepted. This paper aims to give a comprehensive review of the most state of the art brain tumor segmentation methods. We have given a brief introduction to the imaging modalities and their usage in brain tumor segmentation task. We have discussed the results of the most effective approaches by comparing their Dice Score results. We have also discussed some publicly available brain datasets. Furthermore, we have presented a Novel approach for Glioma tumor segmentation using ResNeXt architecture. Experimental results prove that our framework performs well on the dice score.