The AI tool that can predict heart attacks a decade in advance

Every 5 minutes, someone in the UK is admitted to hospital with a heart attack. While survival rates have improved, recovery is often long and challenging for patients and their loved ones. The emotional and psychological scars of heart attacks can also linger long after a patient has physically recovered.

BHF Professor Charalambos Antoniades, based at the University of Oxford, has developed CaRi-Heart®, an AI tool that can help predict someone’s risk of having a heart attack up to 10 years before it happens.

Discovering key warning signs

When someone experiences chest pain that might be due to coronary heart disease (CHD), they often have a computerised tomography (CT) scan of their coronary arteries to check for any narrowed or blocked areas. About three-quarters of the time, the scan finds no significant narrowing of the arteries. But up to two-thirds of these people will go on to have a heart attack in the future.

"We use AI to find the arteries that are more likely to cause a heart attack."

Professor Charalambos Antoniades

Professor Antoniades’ research discovered that inflammation in the fat surrounding the coronary arteries is a key warning sign of a future heart attack. CaRi-Heart® uses AI to detect this inflammation from CT scans and provide a personalised risk score, enabling preventative treatment years before a heart attack strikes. It’s a powerful example of how BHF-funded research is helping to prevent heart attacks before they happen.

This technology has the power to save countless lives and keep people with those they love most for longer.