An artificial intelligence algorithm paired with a smartwatch’s single-lead electrocardiogram (ECG) sensor has shown the ability to detect structural heart diseases in adults, according to a preliminary study to be presented at the American Heart Association's Scientific Sessions 2025. This marks the first prospective study demonstrating that AI can identify multiple types of structural heart disease—such as weakened heart muscle, valve damage, or thickened heart walls—using data from a single-lead ECG captured by a smartwatch. Currently, structural heart conditions are typically diagnosed through echocardiograms, which require specialized equipment and are not widely used for routine screening. The new approach could enable early detection of these often-hidden conditions using devices many people already own. The research team, led by Arya Aminorroaya, M.D., M.P.H., a resident at Yale New Haven Hospital and a research affiliate at the Cardiovascular Data Science (CarDS) Lab at Yale School of Medicine, developed the AI model using over 266,000 12-lead ECG recordings from more than 110,000 adults treated at Yale New Haven Hospital between 2015 and 2023. The algorithm was trained to identify signs of structural heart disease from a single lead of the ECG, simulating the data collected by smartwatch sensors on the back and digital crown of the device. To improve real-world reliability, researchers introduced artificial "noise" into the training data—mimicking the signal interference common in everyday smartwatch use. This helped the AI model become more resilient and accurate when processing imperfect or low-quality signals. The model was first validated using data from 44,591 adults in community hospitals and 3,014 participants from the ELSA-Brasil study, a large population-based health study in Brazil focused on cardiovascular disease and diabetes. In the final phase, the team conducted a prospective study with 600 participants who completed a 30-second single-lead ECG on a smartwatch on the same day they received a heart ultrasound. The average age of participants was 62, with about half being women and diverse racial and ethnic backgrounds. Around 5% of the group had structural heart disease confirmed by ultrasound. The results showed that the AI-powered smartwatch ECG could effectively screen for these conditions, even though a single-lead ECG alone is not sufficient to replace a full 12-lead ECG. The study’s senior author, Rohan Khera, M.D., M.S., director of the CarDS Lab, emphasized that while the single-lead ECG is limited in isolation, AI transforms it into a powerful tool for large-scale, early screening. Limitations of the study include the relatively small number of individuals with confirmed disease in the prospective phase and a rate of false positives. The researchers plan to test the tool in broader, more diverse populations and explore its integration into community-based heart health programs. The goal is to improve preventive care and catch heart disease earlier, potentially reducing the risk of serious complications.