“Imperial College London researchers built an AI model that can detect heart failure and valve disease from a routine ECG in under two seconds, trained on 10.6 million ECGs and tested on 65,000 patients, correctly flagging heart failure in 81% of cases and valve disease in 90%, with a 590-patient trial starting at six hospitals ahead of a…”

The underlying work is real, recent, and from the named team. Artificial intelligence could find signs of heart failure and heart valve disease in a routine ECG, according to research funded by the British Heart Foundation, led by Imperial College London and presented at the European Society of Cardiology Congress in Munich. ESC Congress 2026 took place in Munich and online from 28 to 31 August 2026. On training: to develop the AI, researchers gave it 10.6 million ECGs together with the clinical reports describing them, and the model was then trained to learn the connection between the patterns in the ECG results and specific heart conditions, using 72,475 ECGs linked to scan results revealing the structure and function of the heart. On testing: the technology was tested on the ECGs of a small group of 5,442 patients in the US, and the ECGs of a larger group of 61,520 US patients, whose echocardiogram results could be compared with the AI analysis. The cohort-level breakdown is where the claim's two headline percentages come apart. Researchers tested whether the AI could identify reduced left ventricular ejection fraction, as well as thickening of the heart muscle, and heart valve disease of moderate severity or higher. It correctly identified 77 per cent of the small group of 5,442 patients, and 81 per cent of the large group of 61,520 patients who had poor heart pumping function. For valve disease, the AI was able to identify several forms of heart valve disease, although its performance varied depending on which valve was affected. For aortic stenosis, the AI correctly identified 90 per cent of the small group of more than 5,000 patients, and 80 per cent of the larger group of more than 61,000 patients, who had aortic stenosis. For reduced heart pumping function, the AI achieved an area under the receiver operating characteristic curve of 0.86 for the set of 5,442 patients, and 0.9 for the larger group. For aortic stenosis, it achieved a score of 0.85 and 0.73, respectively. The researchers themselves scope the tool as triage rather than diagnosis. The technology cannot be used on its own to definitively diagnose or rule out heart failure or heart valve disease, but it gives a strong indication that someone may have these conditions from their ECG. Someone judged as high-risk could be sent rapidly for an ultrasound heart scan called an echocardiogram, if this technology became available in the future. On the NHS study: the AI is now being tested on ECGs from 590 NHS patients across London and Bristol, and the team say the AI ECG technology could be two years away from being used routinely for patients. The NHS trial will recruit patients who are being treated at Chelsea and Westminster Hospital, West Middlesex University Hospital and Hammersmith and St Mary's hospitals, in London, as well as Bristol Royal Infirmary and Southmead Hospital in Bristol. That is six named hospitals. The corresponding registry entry, NCT07057466, "Prospective Evaluation of Artificial Intelligence-enhanced Electrocardiography for Detection of Structural Heart Disease" (AI-ECG-SHD), sponsored by Imperial College London, is listed as RECRUITING with an actual start date of 4 November 2025 and an estimated primary completion date of 3 May 2027, and it lists five recruiting UK sites: Southmead Hospital in Bristol, and Chelsea and Westminster, Hammersmith, St Mary's and West Middlesex University hospitals in London, with El-Medany as contact. Context on the team and the commercial vehicle: the work, led by Professor Fu Siong Ng's group, is being taken forward into a spinout company called Cardiovolt.ai.