Towards an Improved CNN Architecture for Brain Tumor Classification

Abstract

Machine learning (a subset of Artificial Intelligence) automatically creates analytic models that adapt to what they find in the data. Over time, the algorithm “learns” how to deliver more accurate results, whether the goal is to make smarter credit decisions, retail offers, medical diagnoses or fraud detection. The use of Deep Learning technology as a new Machine Learning tools has had considerable success in Digital Image Processing over the past few years. It has been widely used in several of complex problems and has proven to be a powerful solving tool. In this paper we present a comparative study between the famous convolutional architectures (LeNet, AlexNet, ZF Net, GoogLeNet, VGGNet, ResNets, DenseNet) and the Convolutional Neural Networks AsilNet that we propose concerning the classification of Brain Tumors (Aneurysms, Multiple Sclerosis, Hydrocephalus, Stroke, Infections, Cysts, Swelling, Hemorrhage, Bleeding, Inflammation) detected in the Magnetic Resonance Imaging.