Fast Diffusion-Relaxation MRI with AI for Prostate Cancer Grading at Ultra-High Gradient Strength

In Wales, around 600 men die from prostate cancer each year. A key step in treating prostate cancer is determining the severity, or grade, of the tumour. The grade helps doctors understand how likely the cancer is to spread, which is essential for choosing the right treatment. Currently, tumour grade is usually determined through needle biopsies. These are invasive and can cause pain, bleeding, and infection. They also rely on the needle hitting the tumour, otherwise results can be misleading.

MRI is emerging as a non-invasive alternative for grading prostate cancer, but current MRI methods are relatively simple and do not always provide enough detail to accurately assess tumour severity. This project will use an advanced MRI technique, called combined diffusion-relaxation MRI, which can provide more accurate tumour grading than standard MRI. The challenge is that this technique can take up to an hour in the scanner and requires complex analysis.

To overcome this, we will use machine learning to design a faster scanning method and improve the accuracy of image analysis. We will test these new techniques on prostate cancer patients and compare the results with traditional biopsy grading to see if our non-invasive approach improves on current methods.

If successful, this project will lay the groundwork for improved non-invasive testing for prostate cancer patients.