Associations between echocardiographic traits and artificial intelligence-enabled electrocardiography predictions of heart failure

Artificial intelligence-enabled electrocardiography (AI–ECG) can detect heart failure (HF), including disease not captured by left ventricular ejection fraction (LVEF), but the cardiac phenotypes underlying AI–ECG HF prediction remain unclear. We therefore investigated whether an AI–ECG HF prediction score aligns with established echocardiographic measures of myocardial dysfunction, remodelling, and filling pressures.

We retrospectively analysed ECG and echocardiography data from 8147 patients who underwent both examinations within 3 days at Akershus University Hospital between 1 January 2023 and 1 June 2025. A previously developed AI–ECG model, pragmatically trained using ICD-10 HF codes and N-terminal pro-B-type natriuretic peptide thresholds, was applied to all electrocardiograms. Spearman’s rank correlation ρ quantified associations between echocardiographic parameters and the AI–ECG HF prediction score. Subgroup analyses were performed by sex and LVEF. External replication of shared echocardiographic associations included 36 286 ECG–echocardiography pairs from Columbia University Irving Medical Center. Global longitudinal strain (GLS) showed the strongest correlation (ρ = 0.57), followed by mitral annular plane systolic excursion (MAPSE; ρ = −0.49) and LVEF (ρ = −0.45). In patients with LVEF ≥50%, correlations remained substantial for GLS, MAPSE, and diastolic-related parameters. Volumetric left ventricular indices correlated less strongly in women, whereas diastolic indices showed stronger correlations in women than in men.

Echocardiographic trait characterization showed that the AI–ECG HF prediction score aligned primarily with measures of systolic function, particularly GLS, while also being associated with diastolic-related abnormalities in patients with preserved LVEF. This approach may inform future studies of model interpretability and refinement.