Brain Tumor Detection Using Convolutional Neural Network

Abstract

Brain tumour is one of the most dangerous disease today. As it spreads rapidly, early detection is needed for the survival of the patient. Diagnosing manually is time consuming and can be erroneous too. Apart from that it needs a group of experts from different specialisations such as pathologists, technologists etc. In the proposed system discussed in this paper we are implementing a method of automatic detection of brain tumour using Convolutional Neural Network (CNN). We are using MRIs as samples. CNN is used to build the model and train the model using past brain tumour patients’ records and using that model to predict a new image if there it is diseased or not. It is difficult to get enough samples to train the model. So we are also applying augmentation to the dataset to increase the number of samples to get better accuracy.