Abstract

Brain tumor is one of the most dangerous and deadly cancer among adults and children. Early diagnosis of brain tumors plays an important role in improving treatment possibilities and increases the survival rate of patients. Magnetic resonance imaging (MRI) is a widely used imaging technique to assess these tumors but the manual segmentation using the MRI images is a time-consuming task. Deep learning methods can enable efficient processing and objective evaluation of the large amount of MRI based image data. The objective of this project is the automatic segmentation of MRI images of brain tumor using deep learning methods with maximum possible accuracy. The three main stages used are preprocessing, feature extraction, classification via convolutional neural network(CNN) and post processing. Training dataset of brain tumor into the machine and developing a code so as to classify its various types is the fundamental of this project. Keywords: Preprocessing, feature extraction, CNN, post processing.