AI Model Predicts 10-Year Stroke Risk from Standard ECG

Stroke remains a major cause of death and long-term disability worldwide. Clinicians need scalable tools to identify people who face sustained risk years before an event occurs. Existing clinical scores can be cumbersome for routine use and may miss subtle cardiac signals linked to stroke. To help address this challenge, researchers have developed an artificial intelligence (AI) model that estimates 10-year stroke risk from a standard electrocardiogram.

Developed by a team co-led by Mass General Brigham (Boston, MA, USA) with collaborators at the Broad Institute of MIT and Harvard (Cambridge, MA, USA), the model is named ECG2Stroke. It analyzes a single 10‑second electrocardiogram (ECG) and incorporates only age and sex to generate a risk estimate. The approach uses deep learning to capture fine-grained waveform patterns from a test that is inexpensive, noninvasive, and already embedded in cardiology workflows.

The investigators trained the model using data from patients at Massachusetts General Hospital. They then evaluated performance in external cohorts from Brigham and Women’s Hospital and Beth Israel Deaconess Medical Center. In total, records from more than 200,000 patients were used to train and validate the system across institutions.

ECG2Stroke predicted the risk of stroke up to 10 years in the future with performance comparable to a validated clinical risk score across hospitals and patient subgroups. Model interpretation indicated that features linked to atrial dysfunction, occurring in the heart’s upper chambers, were among the strongest drivers of predictions. The algorithm was particularly accurate for cardioembolic stroke, which results from blood clots that form in the heart and is preventable with blood thinners. The findings are published in JACC.

“Existing tools to identify which patients are at the highest risk of stroke often require cumbersome clinical score calculations, are not easily scalable, and are therefore not used widely in routine practice,” said Rahul Mahajan, MD, Ph.D., a neurologist with Mass General Brigham Neuroscience Institute and the Broad Cardiovascular Disease Initiative.

“If confirmed after prospective, real-world studies, tools like this could identify which patients should be prioritized for intensive prevention efforts. The tool could also be helpful in driving future mechanistic research into abnormalities in the upper chambers of the heart and links to stroke,” said Shaan Khurshid, MD, MPH, a cardiologist with Mass General Brigham Heart and Vascular Institute and the Broad Cardiovascular Disease Initiative.

Related Links
Mass General Brigham
Broad Institute