A Systematic Review of Deep Learning and Machine Learning Methods in Diagnosis of Brain Tumor

Abstract

Brain tumour diagnosis is a major challenge for the medical professionals now a days. Due to their inexorable growth, brain tumors require early detection to improve patient survival rates. Due to the abundance of important diagnostic information, it offers, magnetic resonance imaging (MRI) has become the de facto medical imaging technology. Brain tumours are detected in MRI images using machine learning and deep learning, two forms of artificial intelligence. This study aims to provide a thorough examination of the categorization methods based on artificial intelligence (AI) that are currently used to forecast brain cancers. (Machine and Deep Learning). The purpose of this study is to identify the best successful methods for identifying brain malignancies using deep learning, machine learning, and optimization. These include CNN, UNET, supervised machine learning, etc. The achievements and failures of the current models are emphasized, along with the potential for future research to increase the patient survival rate. Further, a comparative analysis of various existing brain tumor classification techniques according to the publication year, published journals, techniques, and performance measures is also done in this paper.