Fast MRI in prostate cancer

Speeding up MRI with AI: fast MRI applications in prostate cancer

This collaboration between the UMCG (Groningen, The Netherlands), RadboudUmc (Nijmegen, The Netherlands), University of Twente (Enschede, The Netherlands) and Siemens Healthineers (Erlangen, Germany) focuses on new Artificial Intelligence (AI) techniques for faster and more accurate MRI for diagnostic and interventional purposes in prostate cancer.  

National and European guidelines have recently changed and recommend MRI as first-line diagnostic work up in all patients suspected of prostate cancer. Together with the ever ageing society the number of MRI scans and associated cumulative costs will substantially increase. To make health care affordable for everyone in the future, innovation with the purpose of faster and accurate MRI techniques is necessary.  

With the help of AI new techniques will be investigated that shorten the length of a prostate MRI for diagnostic purposes and during interventions. By using the raw MR image data and examples of previously detected prostate cancers on MRI the normally 40 minute lasting exam can potentially be reduced to 15 minutes. It is estimated that more than 30.000 MRI scans of the prostate are needed annually. Shorter MRI exams will tackle this potential capacity problem and keep our health care affordable.  

This project will deliver novel AI methods for speeding up the acquisition of prostate MRI for the diagnosis of prostate cancer and during MRI guided treatment of prostate cancer. The fundamentals of optimised MR image sequences with fast acquisition, steered towards suspicious areas/instruments during treatment will be layed down.  

Summary
MRI of the prostate is one of the best available methods to detect significant prostate cancer and its use will substantially increase for the next coming years. This project investigates new AI techniques for faster and more accurate MRI of the prostate for diagnostic and interventional purposes.
Technology Readiness Level (TRL)
1 - 3
Time period
48 months
Partners
Logo
Logo
Logo
Logo