A Comprehensive Review on Brain Tumor Detection Using Advanced Learning Algorithms

Abstract

Early detection of brain tumor is of critical importance in medical imagery. It reviews the application of advanced learning algorithms to increase the accuracy and efficiency of brain tumor detection using noninvasive imaging (in particular MRI). In the last couple of years, recent machine learning and deep learning approaches like Convolutional Neural Networks (CNNs), Support Vector Machines (SVMs), and hybrid models have shown an overnight progress by automating the process of extraction, segmentation, and classification of brain tumors. Tumor sub-region identification and feature selection has been improved by using techniques such as Random Forest classifiers, unsupervised clustering, and ensemble methods. Secondly, detection performance has been improved by the integration of handcrafted and automatic features, including texture, shape, and intensity. Further feature selection process is dimensionality reduction based techniques like Principal Component Analysis (PCA) and Information Gain (IG). The focus of this review is on the increasing usefulness of these algorithms to achieve adequate diagnosis and discuss future trends in personalized diagnostics and treatment planning.