AI that can predict heart attacks in use in hospitals

Advanced technology allows doctors to diagnose heart damage from scans before arteries get clogged

ARTIFICIAL intelligence capable of predicting when people may be about to suffer a heart attack is now being used at UK hospitals.

Using advanced technology, doctors are able to extract more patient information from CAT heart scans compared with standard X-rays, such as whether any inflammation is present in the walls of the organ’s arteries.

Called CaRi-Heart, the AI has been designed to act as an “early warning system” to identify when a person’s heart is damaged but shows no obvious signs of failure such as blockages and plaque build-up.

Researchers say the tool allows doctors for the first time to diagnose heart health issues before artery obstruction occurs.

They add that the breakthrough provides an earlier window to identify high-risk people and intervene to lower the likelihood of them having a fatal heart attack.

“All of the inflammation is there, but we have never been able to access it before CaRi-Heart,” Dr Ronak Rajani, a consultant cardiologist at the Harley Street Clinic and professor of cardiovascular imaging at King’s College London, said.

Dr Rajani, alongside John Deanfield, a cardiology professor at University College London, was instrumental in bringing the technology to Britain’s hospitals.

It is available for private patients in Harley Street for £495. Talks are also under way for it to be made available, free of charge, on the NHS.

Scientists say that although plaque build-up is a serious problem, around half of all heart attacks do not occur in fully blocked arteries.

The other half are caused when small pieces of plaque rupture, releasing cholesterol into the bloodstream and causing a clot that can lead to a heart attack.

Experts say current CAT scans are unable to detect signs of rupture.

But Dr Cheerag Shirodaria, the cofounder and CEO of Caristo Diagnostics, which built the technology, said he learned that the clots occurred in areas of the heart where the artery walls were inflamed.

He identified that inflammation led to tiny areas of discolouration in the heart tissue, and he built an artificial intelligence model capable of quantifying these changes.

It means cardiologists are able to use a new metric, called the Fat Attenuation Index (FAI), to identify areas of the heart where signs are suboptimal.

The technology has been tested by the British Heart Foundation which oversaw a trial involving 4,000 patients who had a CAT scan before their results were sent to an Oxford-based laboratory to be analysed by Caristo.

The nine-year project found that people with an abnormal FAI were up to nine times more likely to die of a heart attack in the next nine years than those with normal FAI readings.

It also revealed that one third of patients who received a low-risk diagnosis from a standard scan were actually at much higher risk when studied by CaRi-Heart.

Each of the heart’s three arteries – left anterior descending, left circumflex and right coronary – are given their own FAI score. The lower, the better.

These are then compared with the rest of the population and a patient is given a percentile rating. The 50th percentile is the average, and being ranked higher indicates a higher heart attack risk.

The system also produces a “CaRiRisk” figure which tells the patient the likelihood of them having a fatal cardiac event in the next eight years.

If this figure is alarmingly high, doctors are able to intervene with either medication, such as aspirin or statins, to alleviate some of the inflammation, or encourage lifestyle changes, such as losing weight, cutting out alcohol and being more active.

Dr Shirodaria said: “If we are really going to get a handle on coronary heart disease, which is still the biggest killer, we need to focus much more on prevention. All of our effort is spent when people have disease and it is already established.

“If we can identify earlier when we can modify the risk then ultimately that translates into significant cost savings for the NHS but also, more importantly, significant lives saved.”