Detecção de tumores cerebrais em imagens de ressonância magnética utilizando redes neurais convolucionais.

Abstract

This report details the process of developing and implementing a Convolutional Neural Network using Transfer Learning from the VGG19 model. It includes explanations and references about the evolution of Computer Vision, as well as practical aspects such as image preprocessing, model building, training, inference, and validation. In addition, it discusses the Grad-CAM technique, which provides insight into the algorithm’s decision-making process through heatmap visualization.