Researchers present preliminary work at the American Heart Association’s Scientific Sessions 2026 on whether AI can analyze smartwatch ECG recordings to flag left ventricular systolic dysfunction (LVSD), a condition where the heart’s pumping function is reduced. One report describes an AI system trained to interpret ECG signals measured by smartwatches, suggesting it may be able to identify LVSD, but also notes that performance varies.
Across the coverage, a key context is that results are not consistent. Reports indicate differences between smartwatch brands and across individual ECG recordings. Another outlet highlights a separate Korean study that raises doubts about reliability “in its current state,” reflecting that detection may be more difficult than early optimism suggests. Together, the articles frame the technology as promising but still limited, with uncertainty about general performance and practical readiness.
