PROPERMED: Prostate cancer - Personalized medicine powered by MRI and AI

NTNU Health

PROPERMED: Prostate cancer - Personalized medicine powered by MRI and AI

In prostate cancer, achieving accurate diagnosis, precise patient stratification, and personalized management and treatment is crucial to improve the patient’s quality of life and decrease mortality. However, prostate cancer is the most frequently diagnosed cancer in men and remains the second leading cause of cancer death in males. Consequently, prostate cancer diagnostics and treatment occupy large healthcare resources, and still, we are unable to precisely stratify and provide the patients with the tailored management and treatment they require. Thus, a typical prostate cancer patient faces diagnostic uncertainty, and the risk of under- or over-treatment. 

The current project will take a leap in the right direction for personalized medicine in prostate cancer. The project builds on recent results and decision support tools developed at NTNU and St. Olavs Hospital based on magnetic resonance imaging and artificial intelligence (AI) for detection of the disease (Patent application submitted May 2023). The generic field of AI is rapidly developing, and direct deployment and optimization of novel AI methodology in medical imaging has an immense potential in personalized medicine. However, only a minority of such solutions have yet manifested in routine clinical use. Thus, bringing this potential into viable, sustainable, and useful tools benefitting clinicians and the individual patients calls for continued interdisciplinary research. Our project targets this through a holistic approach, with clear synergies between 1) technical AI development, 2) prospective clinical testing of AI for prostate cancer detection and biopsy targeting 3) and evaluation of its societal impact. 

This project is truly interdisciplinary and enabled by partners situated at three NTNU Faculties and St. Olavs Hospital, together covering expertise in MRI, AI, sociology, health service, urology, radiology, and oncology.  

Image: Our AI-based algorithm exploits the information from MRI (Here represented by anatomical (T2W) and diffusion weighted (ADC) images. The algorithm provides a tumor probability map which can guide where to target biopsies. The tool will be tested in a prospective proof of technology study in 2023.