Identification of Glioma from MR Images Using Convolutional Neural Network

This paper presents a novel approach of classifying the type of glioma using convolutional neural network (CNN) on 2D MR images. Glioma, most common type of malignant brain tumor, and can be classified according to the type of glial cells affected. The types of gliomas are, namely, actrocytoma, oligodendroglioma and glioblastoma multi-forme (GBM). Various image processing and pattern recognition techniques may be used for cancer identification and classification. Though in recent years deep learning has been proved to be efficient in computer aided diagnosis of diseases. Convolutional Neural Networks, a type of deep neural network which is generally used for classification of images, contains multiple sets of conv-pool layers for feature extraction, followed by fully-connected (FC) layers that make use of extracted features for classification.