Heart failure & cardiomyopathiesDiagnostic or prognostic study
Multisite, External Validation of an AI-Enabled ECG Algorithm for Detection of Low Ejection Fraction
- Multisite, External Validation of an AI-Enabled ECG Algorithm
- four geographically diverse sites in the United States
Authors' conclusion
External validation studies such as this one provide a rigorous framework to validate an algorithm's performance. This study demonstrated the algorithm's strong diagnostic accuracy over a geographically diverse, independent set of patients. In this generally unselected population, the algorithm produced a test negative result in 78% of the cases, suggesting potential utility as a rule-out strategy to defer echocardiography when other clinical findings are absent.
Study context
- Population
- Patients with both ECGs and transthoracic echocardiograms performed within 30 days of each other in clinical practiceevaluable set, records of 16,000 subjects extracted, mean age 66 years, 52% male
- Participants
- 13,960 patients
- Comparison
- ECG-AI against transthoracic echocardiogram for LEF detection
Performance
As reported in the abstract
Diagnostic performance for LEF
- AUROC
- 0.92 [95% CI: 0.91-0.93]
- Sensitivity
- 84.5% (95% CI: 82.2%-86.6%)
- Specificity
- 83.6% (95% CI: 82.9%-84.2%)
- Prevalence of LEF
- 7.9%
- LEF among ECG-AI negative subjects
- 1.6%
- LEF among ECG-AI positive subjects
- 30.5%
- Positive predictive value
- 30.5% (95% CI: 28.8%-32.1%)
- Negative predictive value
- 98.4% (95% CI: 98.2%-98.7%)
Original publication
Multisite, External Validation of an AI-Enabled ECG Algorithm for Detection of Low Ejection Fraction. JACC Adv. Published 16 January 2026. doi:10.1016/j.jacadv.2025.102537
Summarised from the published abstract; every number is checked against it