Abstract

Abstract: DeepLearning is the newest and the current trend of the machine learning field thatpaid a lot of the researchers' attention in the recent few years. As a provenpowerful machine learning tool, deep learning was widely used in severalapplications for solving various complex problems that require extremely highaccuracy and sensitivity, particularly in the medical field. In general, thebrain tumor is one of the most common and aggressive malignant tumor diseaseswhich is leading to a very short expected life if it is diagnosed at a highergrade. Based on that, brain tumor classification is a very critical step afterdetecting the tumor in order to achieve an effective treating plan. In thispaper, we used Convolutional Neural Network (CNN) which is one of the mostwidely used deep learning architectures for classifying a dataset of 3064 T1 weightedcontrast-enhanced brain MR images for grading (classifying) the brain tumorsinto three classes (Glioma, Meningioma, and Pituitary Tumor). The proposed CNNclassifier is a powerful tool and its overall performance with an accuracy of98.93% and sensitivity of 98.18% for the cropped lesions, while the results forthe uncropped lesions are 99% accuracy and 98.52% sensitivity and the resultsfor segmented lesion images are 97.62% for accuracy and 97.40% sensitivity.

KEYWORDS: Convolutional neural networks, medical imageanalysis, machine learning and deep learning.