Abstract

Deep Learning is the newest and most popular trend in the machine learning industry, and it has captured the interest of a significant number of academics in the past few years. Deep learning is a strong technique for machine learning that has been widely employed in a variety of applications for the purpose of solving numerous complicated problems that call for an exceptionally high level of accuracy and sensitivity, especially in the field of medicine. If it is identified at a higher grade, a brain tumor is one of the most prevalent as well as one of the most aggressive forms of malignant tumor disease. As a result, the average life expectancy of a person with this disease is significantly shortened. In light of this, the classification of a brain tumor is an extremely important step that must follow the discovery of the tumor in order to formulate an appropriate treatment strategy. In this paper, we used Convolutional Neural Network (CNN), which is one of the most widely used deep learning architectures, to grade (classify) the brain tumors into three classes (Glioma, Meningioma, and Pituitary Tumor). The dataset that we used consisted of 3064 T1 weighted contrast-enhanced brain MR images. The proposed CNN classifier is an effective piece of equipment, as evidenced by its overall performance, which has an accuracy of 98.93% and a sensitivity of 98.18% for the cropped lesions. In comparison, the results for the uncropped lesions are 99% accuracy and 98.52% sensitivity, and the results for segmented lesion images are 97.62% for accuracy and 97.40% sensitivity.

Description