Technology developed using artificial intelligence could identify people at high risk of a fatal heart attack at least 5 years before it strikes, according to research carried out at the University of Oxford.
Currently, when someone goes to hospital with unexplained chest pain they often have a coronary CT angiogram (CCTA) – a scan of the coronary arteries to check for any narrowed or blocked areas. If there is no significant narrowing, which is seen in about 75 per cent of scans, people are sent home.
Read more about AI:
- Artificial intelligence could lead to more accurate breast cancer diagnoses
- The rise of the conscious machines: how far should we take AI?
However, some of them will still have a heart attack at some point in coming years and there are no methods currently in place that can be used by doctors to spot all of the underlying red flags for a future heart attack.
Now, researchers at the University of Oxford have developed a new method of identifying future risk of heart attack using machine learning – an application of artificial intelligence that provides systems the ability to automatically learn and improve from experience without being explicitly programmed.
Dubbed the fat radiomic profile (FRP), the system looks for signs of inflammation, scarring and changes to the blood vessels that supply blood to the heart.
“Just because someone’s scan of their coronary artery shows there’s no narrowing, that does not mean they are safe from a heart attack,” Professor Charalambos Antoniades, BHF Senior Clinical Fellow at the University of Oxford.
“By harnessing the power of AI, we’ve developed a fingerprint to find ‘bad’ characteristics around people’s arteries. This has huge potential to detect the early signs of disease, and to be able to take all preventative steps before a heart attack strikes, ultimately saving lives. We genuinely believe this technology could be saving lives within the next year.”
The team analysed the expression of genes associated with inflammation, scarring and new blood vessel formation in fat biopsies taken from 167 people undergoing cardiac surgery and compared these to CCTA scan images to look for features that indicated changes to the fat surrounding the heart vessel.
They then trained the FRB using data from a pool 5489 patients, 101 of whom went on to have a heart attack within 5 years of having a CCTA scan. In subsequent tests, the FRB outperformed any of the tools currently used in clinical practice. The team now hope to roll out the technology as soon as next year.