A survey of deep learning based brain tumor segmentation methods, By Hojjat Emami

Research

Title	A survey of deep learning based brain tumor segmentation methods
Type	Presentation
Keywords	Brain tumor segmentation, artificial intelligence, deep learning, comparison framework
Researchers	Amir hossein Kargar Khabbazi، Hojjat Emami

Abstract

The brain tumor is referred to the growth of abnormal cancerous or non-cancerous cells in the brain. One of the most important challenges in diagnosing and treatment of tumors is the reliable segmentation of tumors within medical images. Brain tumor segmentation (BTS) is the process of diagnosing, delineating, and separating tumors from normal brain tissue. The objective is to find out the exact boundary of healthy and cancerous tissues by carefully examining the brain cells. BTS methods are broadly divided into three categories: manual, semi-automatic, and fully automatic. One of the important branches of fully automatic BTS methods is deep-learning (DL). To determine how DL-based BTS approaches have evolved in recent years, this paper reviews the literature from 2012 to 2022.