Artificial intelligence algorithms for prostate cancer prediction by breath analysis

Breath analysis is emerging as a promising screening technique. Analysis of volatile organic compounds by artificial intelligence can lead to an early warning bell by noninvasive screening. In this work, we analyze volatile organic compounds from patients who had undergone a prostate cancer exam. We expose a computational methodology to discriminate suspected and full-blown patients. We adopt a sample selection and use oversampling to train a neural network based on a Multi-Layer Perceptron in order to predict real data and simulate performances on further patients undergoing prostate cancer screening.