Wide & Deep neural network model for patch aggregation in CNN-based prostate cancer detection systems

Prostate cancer (PCa) is one of the most commonly diagnosed cancer and one of the leading causes of death among men, with almost 1.41 million new cases and around 375,000 deaths in 2020. Artificial Intelligence algorithms have had a huge impact on medical image analysis, including digital histopathology, where Convolutional Neural Networks (CNNs) are used to provide a fast and accurate diagnosis, supporting experts in this task. In this work, a novel patch aggregation method based on a custom Wide & Deep neural network model is presented, which performs a slide-level classification using the patch-level classes obtained from a CNN. An accuracy of 94.24% and a sensitivity of 98.87% were achieved, proving that the proposed system could aid pathologists by speeding up the screening process and, thus, contribute to the fight against PCa.