Article | Jan 18, 2022

AI echo reports – automated echo diagnosis support for interpretation of systolic and diastolic function published in The Lancet Digital Health

Key Findings
- Developed on 1,145 echocardiograms and validated across 5 external cohorts totalling up to 10,030 studies, this is one of the most extensively validated AI echo pipelines published, covering both systolic and diastolic function across geographically diverse populations in Asia, Europe, and North America.
- In external datasets, mean absolute errors were 9-25 mL for LV volumes and 6-10% for LVEF, and the AI reliably classified systolic dysfunction (LVEF below 40%) with AUC 0.90-0.92 and diastolic dysfunction (E/e' ratio of 13 or above) with AUC 0.91, all with narrow 95% confidence intervals.
- An independent prospective evaluation confirmed all individual equivalence coefficients were below 0 for all measurements, meaning automated variance was lower than expert inter-reader variance, pre-qualifying the algorithm for the IEC framework used in regulatory submissions.
- Automated measurements were comparable or superior to manual measurements across 11 internal consistency metrics, and demonstrated applicability to serial monitoring, supporting use in longitudinal clinical trial settings.
- Published in The Lancet Digital Health, the paper established the foundational evidence for AI-automated echocardiographic interpretation for systolic and diastolic function, with the authors concluding it could accelerate access, improve quality, and reduce costs in diagnosing and managing heart failure globally.
Tromp, J., Seekings, P. J., Hung, C.-L., Iversen, M. B., Frost, M. J., Ouwerkerk, W., Jiang, Z., Eisenhaber, F., Goh, R. S. M., Zhao, H., Huang, W., Ling, L.-H., Sim, D., Cozzone, P., Richards, A. M., Lee, H. K., Solomon, S. D., Lam, C. S. P., & Ezekowitz, J. A. (2021). Automated interpretation of systolic and diastolic function on the echocardiogram: a multicohort study. The Lancet Digital Health, 4(1). https://doi.org/10.1016/S2589-7500(21)00235-1