A Review on Brain Tumor Prediction using Deep Learning

Detection and segmentation of brain tumors is important in the healthcare domain. Since brain tumors can possibly lead to cancer, it is a crucial task to detect it early through Magnetic Resonance Imaging (MRI) or Computed Tomography (CT), which are the techniques that use radio waves and magnetic fields to present a detailed view of the body organs. The images obtained from the MRI makes it hard to locate the exact position of the tumor and hence it is a challenging task to detect the tumor accurately. Thus, computer-aided methods (segmentation, detection and classification processes) with better accuracy are required for early tumor diagnosis. The segmentation of brain tumor which is usually carried out manually by the radiologists through their expertise and skill is a highly prolonged task and there can be chances of some faulty predictions, hence, the semantic segmentation is proven to be an effective method to overcome this problem. Semantic segmentation method is applied to brain tumors which are automatically segmented with the aid of deep learning techniques (CNN, RNN, GAN, LSTMs, etc.). The usage of deep learning techniques with greater accuracy and robustness are proven to be effective for the precise diagnosis of brain tumor. The primary objective of this paper is to examine the previously published techniques using deep learning for the human brain tumor prediction.