Classification of Brain Tumor and its types using Convolutional Neural Network

A brain tumor is a cancerous (malignant) or non- cancerous (benign) abnormal growth of cells or mass in the brain. The identification of the tumor is done by magnetic resonance imaging (MRI) or a CT scan. The timely detection of brain tumors plays a vital role as it can be life-threatening if left untreated. The traditional method of manually checking the MR image might not be very accurate since it is dependent on the skill of the person examining the images. To increase the efficiency and accuracy of diagnoses by the radiologists and neurologists, we propose a model which uses Convolutional Neural Networks (CNN) based on deep learning techniques to classify the common types of benign tumor. The dataset consists of MRI images of three different labelled brain tumors that are commonly found: Meningiomas, Gliomas and Pituitary Adenomas. The proposed model is first trained using a large number of labelled images and then the model classifies any given MRI image into one of the three above mentioned classes.