Machine Learning Based Intelligent Diagnosis of Brain Tumor: Advances and Challenges

One of the fatal diseases that kills a large number of humanity across the globe is brain tumor. If the brain tumor detection is delayed, then the patient has to spend a large amount of money as well as to face severe suffering. Therefore, there is an essential need to detect the brain tumor so that money and life can be saved. The conventional examination of brain images by doctors does not reveal the presence of a tumor in a reliable and accurate manner. To overcome these issues, early and accurate brain tumor identification is of prime importance. A short while ago, methods employing machine learning (ML) and artificial intelligence (AI) were utilized to properly diagnose other diseases using test attributes, electrocardiogram (ECG), electromyography (EMG), Heart Sounds, and other types of signals obtained from the human body. This chapter presents a complete overview of the detection of patient-provided brain MR pictures and classifying patients’ brain tumor using AI and ML approaches. For this pose, brain images obtained from kaggle.com website have been employed for developing various AI and ML classifiers. Through simulation-based experiments conducted on the AI and ML classifiers, performance matrices have been obtained and compared. From the analysis of results reported in the different articles, it is observed that Random Forest exhibit superior detection of brain tumor. There is still further scope for improving the performances as well as developing affordable, reliable, and robust AI-based brain tumor classifiers.