Medical need: For a considerable number of men with suspected PCa, biopsy of the prostate causes physical stress, pain, and potentially life-threatening adverse consequences. In parallel, demands of cancer care in Europe continually increases, with the number of incident cancer cases in Europe projected to increase by more than 14.1% by 2030. This leads to a growing demand for innovative and non-invasive diagnostic approaches among taxpayers and society. In the era of personalized oncology, in which understanding of the complex biology of cancer is a prerequisite, a reliable diagnostic setup is warranted to obtain all information needed. Current therapeutic approaches for primary prostate cancer patients are based on the gold-standard visual grading of the biopsy slides (Gleason score or ISUP grade). Based on this prognostic information, patients can be assorted into PCa risk groups to define the respective treatment options. However, the Gleason score is not a predictive marker and thus the individual patient’s treatment response to targeted drugs or radiation therapy cannot be predicted. 

Hypothesis: Prostate cancer biopsy can be replaced by deep-learning based interpretation of advanced medical imaging, achieving a non-invasive prostate cancer tissue characterization.
‍
Outcome 1: We will create a deep-learning based model (software tool) for non-invasive prostate cancer characterization in terms of Gleason score and tumor stiffness.
‍
Outcome 2: We will create 3D model of tumor stiffness from MR-elastography for advanced mathematical modeling of drug delivery and radiobiological modeling of the effect of radiotherapy for prostate cancer.
‍
Partners: German Oncology Center, The Cyprus Institute, University of Cyprus, IEO Milano (Italy), University of North Carolina – Chapel Hill (USA)