The aim of this study was to develop and validate on multivendor data a fully automated computer-aided diagnosis (CAD) system based on artificial intelligence, to automatically localize, segment, and classify PCa lesions according to their aggressiveness.