A Critical Evaluation of Deep Learning Methods for Classifying Brain Tumor: A Review

According to a February 2018 WHO report, Asia has the highest brain or CNS cancer death rate. Early tumor detection can save many of these lives. Targeted therapy requires tumor classification. Due to the high cost, lengthy process, and risk of infection associated with diagnosing a tumor, we have an immediate need for non-invasive, cost-effective, and efficient methods of describing and determining the grade of brain tumors. MRI, CT, and other brain scans can detect tumors quickly and safely. This study summarizes brain cancer pathophysiology, imaging modalities, and machine and deep learning-based brain tumor characterization.