An artificial intelligence algorithm paired with single-lead electrocardiogram sensors on smartwatches accurately diagnosed structural heart diseases such as weakened pumping ability, damaged valves, or thickened heart muscle, according to a preliminary study to be presented at the American Heart Association's Scientific Sessions 2025. This research represents the first prospective study demonstrating that AI can detect multiple structural heart diseases using measurements from smartwatch sensors.
Millions of people wear smartwatches that are currently mainly used to detect heart rhythm problems such as atrial fibrillation. Structural heart diseases, however, are typically identified through echocardiograms, advanced ultrasound imaging tests requiring specialized equipment not widely available for routine screening. The study explored whether everyday smartwatches could help detect these hidden structural heart conditions earlier, before they progress to serious complications or cardiac events.
Researchers developed the AI algorithm using more than 266,000 12-lead ECG recordings from over 110,000 adults. Based on this extensive data library, they created an algorithm to identify structural heart disease from single-lead ECGs obtainable through smartwatch sensors. The team isolated one of the 12 ECG leads to resemble smartwatch single-lead ECGs and incorporated random interference or noise that could occur during real-world smartwatch recordings to enhance the model's resilience.
The AI model underwent external validation using data from community hospital patients and participants from the population-based ELSA-Brasil study. In the prospective real-world evaluation, 600 participants underwent 30-second single-lead ECGs using smartwatches on the same day they received heart ultrasounds. The analysis revealed the AI model scored 92% on standard performance metrics using hospital equipment single-lead ECGs and maintained 88% performance with smartwatch-obtained ECGs. The algorithm demonstrated 86% sensitivity for identifying people with heart disease and 99% negative predictive value for accurately ruling out heart disease.
Study author Arya Aminorroaya, M.D., M.P.H., emphasized that while a single-lead ECG alone cannot replace the 12-lead ECG tests available in healthcare settings, combining it with AI creates powerful screening capabilities. Senior author Rohan Khera, M.D., M.S., noted this approach could enable large-scale early screening for structural heart disease using devices many people already own. The study population had a median age of 62 years, with approximately half being women and diverse racial and ethnic representation.
Study limitations include the small number of patients with actual disease in the prospective study and the presence of false positive results. Researchers plan to evaluate the AI tool in broader settings and explore integration into community-based heart disease screening programs to assess potential impact on improving preventive care. The findings remain preliminary until published as full manuscripts in peer-reviewed scientific journals, as abstracts presented at American Heart Association scientific meetings are not peer-reviewed.
