AI biomarker predicts heart attacks
5 Sep 2019
A new AI-powered imaging biomarker is said to predict hearts attacks at least five years before they are due to happen.
A University of Oxford team created a fat radiomic profile (FRP), which uses machine learning to detect biological red flags in the perivascular space lining blood vessels to the heart, such as inflammation, scarring and other changes.
The researchers say FRP can predict heart attacks more effectively than tools currently being used in clinical practice, and may also lead to more personalised heart care.
Charalambos Antoniades, Professor of Cardiovascular Medicine and BHF Senior Clinical Fellow at the University of Oxford, said: "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.”
There are no methods used routinely by doctors that can spot underlying red flags for a future heart attack, and that even if a coronary CT angiograms (CCTA) scan does not detect narrowed coronary arteries, it does not mean the patient is safe from a heart attack.
To build the biomarker, the research team used fat biopsies from 167 people undergoing cardiac surgery and analysed gene expression associated with inflammation, scarring and new blood vessel formation. They then matched these to CCTA scan images to determine which features best indicate changes to fat surrounding heart vessels.
They then compared the CCTA scans of the 101 people who went on to have a heart attack or cardiovascular death within five years of having a CCTA with matched controls who did not, to understand the changes in the perivascular space.
As with any machine learning model, the more heart scans that are added, the more accurate predictions will become.
The research was funded by the British Heart Foundation and the National Institute for Health Research. It has been published in the European Heart Journal.
