Using AI to spot aggressive prostate cancer early

Grant information

Reference: MRC-CRTF24-001
Lead researcher: Mr Thineskrishna Anbarasan
Institution: University of Oxford
Award: £245,732 (funded jointly with the Medical Research Council)

What you need to know

Prostate cancer is common, but it’s hard to tell which cases are aggressive and need major treatment.
This project combines MRI scans with genetic maps of the prostate, using artificial intelligence to spot patterns linked to dangerous cancers.
The goal is to create a tool that helps doctors target biopsies and treatments more precisely - reducing unnecessary surgery and side effects while catching high-risk disease earlier.

About Mr Thineskrishna Anbarasan

Mr Thineskrishna Anbarasan completed his medical studies along with an intercalated bachelor’s degree in genetics, cancer and personalised medicine from the University of Dundee. He is currently a Urology resident at the University of Oxford. His research focuses on understanding how prostate cancer behaves and using new technologies to improve diagnosis.

For this project, he’ll be guided by some of the UK’s top experts. Professor Ian Mills brings decades of experience in prostate biology and mentoring clinical academics, while Associate Professor Alastair Lamb specialises in understanding the biology of the disease. Associate Professor Bartolomiej Papiez brings expertise in artificial intelligence and medical imaging, and Professor Richard Bryant is an experienced urologist who knows the challenges men face in real-world care. Together, they’ll help Mr Anbarasan turn cutting-edge science into practical benefits for men.

Why are we funding this research?

Prostate cancer is the most common cancer in the UK, but not all cases are equally dangerous. The challenge is spotting the aggressive cancers - the ones likely to spread - without putting men through unnecessary surgery or harsh treatments they don’t need. At the moment, doctors rely on MRI scans and biopsy samples, but these only give part of the picture. They can miss hidden clues about how risky a cancer really is.

This research takes a fresh approach. It combines MRI images with detailed genetic maps of the prostate, using artificial intelligence to link what’s seen on the scan with what’s happening inside the cells. In doing this, the research aims to uncover patterns that reveal which cancers are truly aggressive.

This could lead to a new tool that helps doctors target biopsies more accurately and choose treatments that fit each man’s risk. In short, it’s about making prostate cancer care smarter - catching dangerous disease early while avoiding unnecessary side effects for those with low-risk cancer.

What will Mr Anbarasan do?

Mr Anbarasan's goal is to connect what doctors see on an MRI with what’s happening inside the tumour at a genetic level. To do this, he’ll start with prostate samples that have already been removed during surgery. Using an advanced technique called spatial genomics, he’ll create detailed maps showing which genes are active in different parts of the prostate.

Next, he’ll take MRI scans of the same prostates and use artificial intelligence (AI) to match these images to the genetic maps. This is a complex task because prostate samples change shape after surgery, so he’ll train computer models to recognise and correct for these differences. Once the scans and genetic maps are aligned, he’ll look for patterns - specific features on the MRI that consistently link to genetic changes known to make cancer more aggressive.

From these patterns, Mr Anbarasan will develop a kind of digital fingerprint that predicts how dangerous a cancer is just from the scan. He’ll then test this score on large groups of men who’ve had MRI scans and long-term follow-up, checking whether it accurately predicts which cancers spread or return.

If successful, his work could transform prostate cancer care - helping doctors target biopsies more precisely and choose treatments that fit each man’s risk, while avoiding unnecessary surgery and side effects for those with low-risk disease.

By better identifying which cancers are truly dangerous, we can target treatment more precisely and avoid unnecessary harm. Receiving this award from Prostate Cancer UK gives me the opportunity to train with world-leading researchers and bring new ideas that could impact how we detect aggressive prostate cancer.

Mr Thineskrishna Anbarasan University of Oxford

How will this benefit men?

This research could make prostate cancer care much more precise.

Right now, MRI scans can show suspicious areas, but they don’t always reveal how aggressive a cancer really is. That means some men have surgery or radiotherapy they might not need, while others with dangerous cancers could be missed.

By linking MRI images to genetic clues inside the tumour, this project aims to create a tool that predicts which cancers are high-risk. For men, that means fewer unnecessary treatments and side effects, and faster, more targeted care for those who need it most.

In short, it’s about giving every man the right treatment at the right time.