Classification and Segmentation of Brain MR Images Using Deep Neural Networks and Fuzzy C-Mean Clustering

Brain tumor arises mainly due to the tissue's abnormal growth inside the brain. They are dangerous as compared to other kinds of tumors because of their aggressively fast-spreading nature leading to a lower survival rate in the patients. Therefore, detecting and segmenting brain tumors from brain MR images helps radiologists in the early treatment planning of patients. The proposed work uses deep neural networks to classify and a fuzzy C-mean clustering algorithm to segment. First, classification is done to detect whether the image is normal or has a tumor. Next, segmentation is done for the MR images with tumors. The evaluated results are also compared with the recently published work.