Advanced electrocardiography heart age: a prognostic, explainable machine learning approach applicable to sinus and...

Area of Science:

• Cardiology
• Medical Technology
• Predictive Analytics

Background:

• The explainable advanced electrocardiography (A-ECG) Heart Age gap estimates cardiovascular aging and risk.
• Existing A-ECG Heart Age methods are limited to patients in sinus rhythm.
• A need exists for a cardiovascular risk assessment tool applicable across all heart rhythms.

Purpose of the Study:

• To develop and validate a revised, explainable A-ECG Heart Age applicable to both sinus and non-sinus rhythms.
• To assess the prognostic value of this new A-ECG Heart Age gap in predicting cardiovascular events and comorbidities.

Main Methods:

• A novel A-ECG Heart Age, excluding P-wave measures, was derived using machine learning on 10-second 12-lead ECGs.
• The model was trained using Bayesian 5-minute A-ECG Heart Age as a reference in a derivation cohort (n=2771).
• External validation in a separate cohort (n=731) assessed associations with cardiovascular events (heart failure hospitalization or death) and comorbidities using Cox regression.

Main Results:

• The derived A-ECG Heart Age showed strong agreement with the reference 5-minute A-ECG Heart Age (R²=0.91).
• In the validation cohort, an increased A-ECG Heart Age gap (≥10 years) was significantly associated with adverse cardiovascular events (HR 2.04, C-statistic 0.58).
• The A-ECG Heart Age gap also correlated with the presence of hypertension, diabetes mellitus, hypercholesterolemia, and heart failure (P ≤ 0.009).

Conclusions:

• An explainable A-ECG Heart Age gap can be reliably calculated irrespective of heart rhythm.
• This revised A-ECG Heart Age gap is a significant predictor of cardiovascular risk, morbidity, and mortality.
• The method offers a valuable tool for intuitive cardiovascular risk communication and clinical decision-making.