CLASSIFICATION OF BRAIN TUMOR FROM MAGNETIC RESONANCE IMAGING USING CONVOLUTIONAL NEURAL NETWORKS
Abstract
Deep learning methods gained a huge popularity in segmentation and classification of medical imaging. In this paper we propose a Convolutional Neural Network (CNN) approach which is one of the top performing methods while also being extremely computationally efficient, a balance that existing methods have struggled to achieve, we use this method as a process for segmenting brain tumor regions from magnetic resonance imaging (MRI) using CNNs. The main task for this method is using a public dataset containing 3,064 T1-weighted contrast enhanced MRI (CEMRI) with different abnormalities from different planes. This novel method of training neural networks on this dataset has proved to be efficient than well-known methods.