AI Stethoscope Spots Heart Valve Disease Earlier Than GPs

Valvular heart disease affects more than half of people over 65, yet it often goes undiagnosed until symptoms become severe. In advanced stages, untreated cases can carry a mortality risk of up to 80% within two years. Current diagnosis relies on echocardiography, which is costly and not suitable for widespread screening, while traditional stethoscope exams frequently miss cases. Now, a new study suggests that artificial intelligence (AI) could detect serious valve disease years earlier by analyzing heart sounds, offering a scalable screening approach.

Researchers at the University of Cambridge, UK) developed an AI algorithm trained on heart sound recordings from 1,767 patients who also underwent echocardiography, the gold standard reference test. Using digital stethoscopes, the team captured short audio samples and trained the algorithm directly on echocardiogram results rather than on heart murmurs alone. This approach enabled the system to identify subtle acoustic patterns associated with valve disease, even in cases without obvious murmurs.

The AI correctly identified 98% of patients with severe aortic stenosis and 94% of those with severe mitral regurgitation. When compared against 14 general practitioners who reviewed the same recordings, the algorithm consistently outperformed every clinician. The results, published in npj Cardiovascular Health, showed that the system minimized false alarms while maintaining high accuracy, particularly for severe disease. The findings indicate that AI-assisted auscultation could provide reliable detection in primary care settings.

The technology is designed as a rapid screening tool to help doctors decide which patients require referral for echocardiography. Only a few seconds of heart sound recording are needed, and minimal training is required to operate the device. Researchers emphasize that further trials in real-world GP settings are necessary before widespread rollout, particularly to improve detection of moderate disease. If validated, the tool could ease pressure on healthcare systems and enable earlier intervention before irreversible heart damage occurs.

“Valve disease is treatable. We can repair or replace damaged valves and give people many more years of healthy life. But timing is everything,” said Professor Rick Steeds, co-author of the study. “Simple, scalable screening tools like this could make a real difference by finding patients before irreversible damage occurs.”

Related Links:
University of Cambridge