AI tool can spot heart disease in under two seconds, study finds
Researchers have developed an AI tool that can detect signs of heart failure and heart valve disease from a routine electrocardiogram (ECG) in less than two seconds, according to findings presented at the European Society of Cardiology congress in Munich. The technology, described as "superhuman" by its developers, is set to be trialled on NHS patients in London and Bristol and could be rolled out nationwide within two years.
The AI model, developed by experts at Imperial College London, works by extracting more information from an ECG than the human eye can typically see. A standard ECG records the electrical activity of the heart and is commonly used to diagnose heart attacks and abnormal heart rhythms, but it cannot detect heart disease on its own. That typically requires an echocardiogram, an ultrasound scan of the heart, which patients often wait months for.
In a study funded by the British Heart Foundation (BHF), the AI tool was tested on US patients and correctly identified up to 81% of those with heart failure and 90% of those with heart valve disease. The tool cannot definitively diagnose on its own, but it gives a strong indication that further testing may be needed, which could help fast-track high-risk patients for urgent echocardiograms.
Dr Sonya Babu-Narayan, clinical director at the BHF and consultant cardiologist, said: "It is exciting to see that AI can now deliver a read-out from an ECG in what feels like the blink of an eye." She added that the technology would not detect everyone with a heart condition, but could help "fast-track" those most likely to have an abnormality.
Prof Fu Siong Ng, senior investigator from Imperial College London, noted that patients can often wait several months for a heart ultrasound scan after referral. "This makes it exciting that our technology could identify patients most at risk of heart failure and heart valve disease, so they could be prioritised for scans faster and more efficiently," he said.
Dr Ahmed El-Medany, a BHF clinical research fellow who led the ECG analysis, said: "These results suggest there is potentially far more information hidden within a routine ECG than we can recognise by looking at it ourselves." He described the tool as a "superhuman AI" and said the important next question is how it performs in NHS patients.
The NHS trial will recruit patients from Chelsea and Westminster Hospital, West Middlesex University Hospital, Hammersmith and St Mary's hospitals in London, and Bristol Royal Infirmary and Southmead Hospital in Bristol. The AI ECGs could become routine in two years, according to the team.
ECGs are one of the most commonly administered tests worldwide, with around one billion undertaken each year. The researchers said the AI model could be used "opportunistically" to catch disease in people for whom the conditions were not suspected, and that it was particularly good at detecting reduced heart pumping function and several forms of heart valve disease.
About one million people in the UK have been diagnosed with heart failure, while 400,000 are unknowingly living with it, according to the BHF. Previous research suggests as many as one in four healthy people aged 60 or above are living with heart valve disease in Britain. Almost 400,000 people in England are on the NHS cardiology waiting list, as reported by com.
Separate AI model diagnoses conditions from facial videos
In a separate study presented at the same congress, researchers from the University of Tokyo and Institute of Science Tokyo showed that AI software could diagnose high blood pressure and type 2 diabetes from five-second videos of faces. The software analysed skin blood-flow patterns and colouring, detecting hypertension with more than 90% accuracy and diabetes with more than 80% accuracy, according to the Daily Mirror.
The research comes as around 16 million people in the UK have hypertension, four million have type 2 diabetes, and over eight million have heart disease, the Daily Mirror reported. Dr Roy Jogiya, chief medical researcher at Heart Research UK, said AI screening could be transformative for the NHS, where patients currently face months of delays.