Convolutional Neural Networks for brain tumor images classification

Abstract

Lately, the brain tumor is taken into account as one of the most detrimental neurological disorders. Therefore, diagnosing precisely and immediately what type of brain tumor disease in order for doctors to give patients plausible treatments in time is of utmost importance. Nevertheless, determining what kind of brain tumor is excessively relied on human factors such as emotion could lead to either unconstant or incorrect performance. Fortunately, by implementing deep learning into the medical background, the aforementioned issue is addressed almost thoroughly. In order to enhance the performance of Convolutional Neural Networks, the paper proposes a myriad of combined architectures between based CNNs, BatchNormalizaiton, and Dropout along with adjusted parameters and methods. Eventual experimental results prove that the proposed approaches outperform the state-of-the-art paper works on the same benchmark brain tumor database.