DescriptionAI-based solutions start impacting our daily life in different aspects including facilitating medical imaging analysis and clinical diagnosis. In the area of diagnosing prostate cancer (PCa), which currently affects one in eight men, there is no well-accepted screening method for this important public health problem. In this presentation, we present an infrastructure that implement high performance computing resource to train and validate a proposed deep learning architecture to predict/detect the loci of prostate cancer by using a multi-parametric MRI (mp-MRI) prostate cancer dataset. The outcomes of the optimized deep learning model are further integrated into the web interface that can assist the radiologists in early prediction of PCa. Our tool has the potential to revolutionize the prostate cancer diagnosis and improve patient quality of life, while decreasing its cost by minimizing unnecessary prostate biopsies and surgeries.