Artificial Intelligence-Based Classification of Brain Tumors

The most difficult aspect of diagnosing a brain tumor is the tumor’s location, size, and shape fluctuations. The diagnosis of a brain tumor depends on the nature and location of the tumor so that doctors can forecast the patient’s survival chances and make treatment selections ranging from surgery to radiotherapy and chemotherapy In addition, pathologists, who are highly trained in the field of neurosurgery, visually examine patients to diagnose brain tumors in clinics. Nonetheless, this process is performed manually, which is not only time-consuming but also laborious and prone to human error. Therefore, the need for a model that could accurately classify brain tumors is crucial. The goal of this chapter is to develop and implement a transfer learning approach for the artificial intelligence-based classification of brain tumors using MRI images by: (i) pre-processing magnetic resonance scans of brain tumors according to their class (glioma, meningioma, or pituitary tumors) (ii) develop a transfer learning model capable of accurately classifying the various types of brain tumors (iii) develop an easy-to-use web application/GUI based on the trained model.