AI Identifies Hidden Heart Valve Defects from Patient’s ECG

Heart valve diseases, affecting over 41 million people globally, can lead to heart failure, hospitalization, and even death. Early diagnosis is critical, yet symptoms like shortness of breath or dizziness are often misattributed, and some patients show no signs until the disease is advanced. Subtle changes in heart function, especially in its electrical activity, often go undetected until it’s too late. Researchers have now created an artificial intelligence (AI) tool to identify risk much earlier, using just a standard electrocardiogram (ECG).

Developed by researchers at Imperial College London (London, UK), the AI model uses a patient’s ECG to predict the risk of developing regurgitant valvular heart diseases. These conditions, which affect the mitral, tricuspid, or aortic valves, involve blood leaking backwards through the heart. The AI detects early structural changes in the heart that are not apparent to doctors, making it possible to flag high-risk patients earlier than ever before physical symptoms or damage appear.

The research team trained the algorithm on nearly one million ECG and echocardiogram records from over 400,000 patients in China. To ensure accuracy across populations, the tool was then validated on over 34,000 patients in the US, showing that it works well across ethnically diverse populations and healthcare systems. The study, published in The European Heart Journal, shows that the AI model successfully predicted valve disease risk 69–79% of the time, showing reliable performance across diverse ethnic groups and healthcare settings.

Additionally, high-risk patients identified by the AI were up to 10 times more likely to develop valve leakage, offering a critical window for prevention. This technology could revolutionize care by identifying patients in need of monitoring or early intervention, long before heart valve disease causes harm. The research follows on from the team’s development of the related AI-ECG risk estimation model, known as AIRE, which can predict patients’ risk of developing and worsening disease from an ECG.

“Our work is harnessing AI to detect subtle changes at the earliest stage from a simple and common test, and we think this could be really transformative for doctors and patients,” said Dr. Arunashis Sau, one of the study leads. "Rather than waiting for symptoms or relying only on expensive and time-consuming imaging tests, we could use AI-enhanced ECGs to spot those most at risk earlier than ever before."