Artificial intelligence-estimated biological heart age using a 12-lead electrocardiogram predicts mortality and

Yong-Soo Baek1,2,3, Dong-Ho Lee2, Yoonsu Jo2

  • 1Division of Cardiology, Department of Internal Medicine, Inha University College of Medicine and Inha University Hospital, Incheon, South Korea.

Insights

Artificial intelligence-estimated heart age from ECGs predicts mortality and cardiovascular events. An AI ECG-heart age significantly older than chronological age increases risks, while a younger age shows inverse effects.

Area of Science:

  • Cardiology
  • Artificial Intelligence
  • Biomedical Engineering

Background:

  • Limited data exists on artificial intelligence-estimated biological heart age (AI ECG-heart age) for predicting cardiovascular outcomes, differentiating it from chronological age (CA).
  • A deep learning algorithm was developed to estimate AI ECG-heart age using standard 12-lead ECGs.
  • The study aimed to evaluate the predictive capability of AI ECG-heart age for mortality and cardiovascular outcomes.

Purpose of the Study:

  • To develop and validate a deep learning model for estimating biological heart age from ECGs.
  • To assess the association between AI ECG-heart age and all-cause mortality.
  • To investigate the relationship between AI ECG-heart age and major adverse cardiovascular events (MACEs).

Main Methods:

  • A deep neural network was trained and validated on 425,051 12-lead ECGs (2006-2021).
  • A holdout test set of 97,058 ECGs was used for validation.
  • Cox proportional hazards models were employed to analyze outcomes after adjusting for comorbidities.

Main Results:

  • The AI ECG-heart age algorithm demonstrated a mean absolute error of 5.8 ± 3.9 years.
  • Patients with AI ECG-heart age 6 years older than CA had significantly higher all-cause mortality (HR 1.60) and MACEs (HR 1.91).
  • Conversely, patients with AI ECG-heart age younger than CA showed reduced mortality (HR 0.82) and MACEs (HR 0.78).
  • ECG features like PR interval, QRS duration, and QT/QTc intervals showed alterations with increasing AI ECG-heart age.

Conclusions:

  • AI-estimated biological heart age significantly impacts mortality and MACEs.
  • AI ECG-heart age serves as a valuable tool for primary prevention and cardiovascular healthcare.
  • The findings highlight the potential of AI in assessing cardiovascular risk beyond chronological age.
Abstract

Related Concept Videos