According to a new study supported by the British Heart Foundation (BHF), technology created using artificial intelligence (AI) could recognize people at high risk of a fatal heart attack at least five years before it happens.
The study outcomes are being presented at the European Society of Cardiology (ESC) Congress in Paris and reported in the European Heart Journal.
Scientists from the University of Oxford have used machine learning to create a new biomarker, or “fingerprint,” known as the fat radiomic profile (FRP). The fingerprint identifies biological red flags in the perivascular space lining blood vessels that supply blood to the heart. It detects scarring, inflammation, and changes to these blood vessels, which are all pointers to an impending heart attack.
When a person approaches a hospital with chest pain, a standard care component is to perform a coronary CT angiogram (CCTA)— a scan of the coronary arteries to look for any blocked or narrowed segments. In case there is no evident narrowing of the artery, which leads to nearly 75% of scans, people are sent home.
However, some of them will still suffer a heart attack at some time in the future. Doctors do not use any routine methods with the ability to detect all of the underlying red flags for an impending heart attack.
In this research, first, Professor Charalambos Antoniades and his colleagues used fat biopsies of 167 people who underwent cardiac surgery. They studied the expression of genes related to inflammation, scarring, and the formation of new blood vessel. Then, these were matched with the CCTA scan images to identify which features best point out changes to the fat that surrounds the heart vessels, known as perivascular fat.
The researchers then compared the CCTA scans of 101 people, from a cohort of 5487 patients, who suffered a heart attack or cardiovascular death within five years of having a CCTA with analogous controls who did not. This was done to perceive the variations in the perivascular space indicating someone is at higher risk of a heart attack.
They used machine learning to develop the FRP fingerprint that picks up the level of risk. The accuracy of the predictions will increase with the number of heart scans added, and will increase the information that will turn into “core knowledge.”
The performance of this perivascular fingerprint was tested in 1,575 people in the SCOT-HEART trial, indicating that the FRP had a remarkable value in predicting heart attacks, more than what can be realized with any of the tools used at present in clinical practice.
The researchers believe that this robust technology will allow a greater number of people to prevent a heart attack, and intend to supply it to healthcare professionals in 2020, hoping that it will be added to regular NHS practice together with CCTA scans within the next two years.
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. 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.
Charalambos Antoniades, Professor of Cardiovascular Medicine and BHF Senior Clinical Fellow, University of Oxford
Antoniades continued, “We genuinely believe this technology could be saving lives within the next year.”
Every 5 minutes, someone is admitted to a UK hospital due to a heart attack. This research is a powerful example of how innovative use of machine learning technology has the potential to revolutionise how we identify people at risk of a heart attack and prevent them from happening.
Metin Avkiran, Professor, Associate Medical Director, British Heart Foundation
Avkiran added, “This is a significant advance. The new ‘fingerprint’ extracts additional information about underlying biology from scans used routinely to detect narrowed arteries. Such AI-based technology to predict an impending heart attack with greater precision could represent a big step forward in personalised care for people with suspected coronary artery disease.”
Apart from the BHF, this study was funded by the National Institute for Health Research (NIHR).