MR Brain Tumour Classification Using a Deep Ensemble Learning Technique

Abstract

A hereditary disease known as a brain tumour can manifest as an odd mass of tissue where cells proliferate and develop, but it is uncontrollable. Magnetic resonance imaging (MRI), which is particularly helpful for visualizing the brain, can be used to find these tumours. However, manual detection takes more time, which results in occasional inaccuracies. In this study, we suggested using an ensemble learning technique to classify brain tumours based on MRI scans. Some deep learning algorithms carry out the task of scouring a hypothesis space in search of an appropriate hypothesis that will produce accurate predictions for a specific tumour situation. To provide a more accurate hypothesis for the prediction of brain tumours, ensembles combine several competing ideas. By putting the combination of pretrained models to the test, a proposed model is developed on top of earlier research on ensemble approaches. In this paradigm, certain ML classification algorithms combine and segment features extracted using transfer learning to enhance performance. With the ensemble-based classifier, we were able to achieve 99% accuracy.