Time-Dependent ECG-AI Prediction of Fatal Coronary Heart Disease: A Retrospective Study
Liam Butler1, Alexander Ivanov1, Turgay Celik1
1Cardiovascular Section, Department of Internal Medicine, Wake Forest University School of Medicine, Winston-Salem, NC 27101, USA.
Insights
Predicting fatal coronary heart disease (FCHD) risk is possible using electrocardiogram artificial intelligence (ECG-AI) models. Even single-lead ECGs combined with demographics accurately forecast 2-year FCHD risk.
Area of Science:
- Cardiology
- Artificial Intelligence
- Medical Informatics
Background:
- Fatal coronary heart disease (FCHD) is a significant cause of mortality in the US.
- Electrocardiogram artificial intelligence (ECG-AI) shows promise for predicting adverse coronary events.
- The application of ECG-AI specifically for FCHD risk prediction remains understudied.
Purpose of the Study:
- To develop and validate ECG-AI models for predicting the risk of fatal coronary heart disease.
- To assess the efficacy of both 12-lead and single-lead ECGs in FCHD risk prediction.
- To evaluate the added value of demographic and clinical data when combined with ECG-AI outputs.
Main Methods:
- Retrospective analysis of 10-second 12-lead ECGs and clinical data from two large cohorts (UTHSC and AHWFB).
- Development of convolutional neural network (CNN) models using 12-lead and Lead I ECGs.
- Integration of ECG-AI outputs with demographic/clinical data using time-dependent Cox proportional hazard models for risk prediction.
Main Results:
- The developed ECG-AI models demonstrated strong predictive performance, with validation AUCs of 0.84 (12-lead) and 0.85 (Lead I).
- The optimal model, combining simple demographics with Lead I ECG-AI output (D1-ECG-AI-Cox), achieved an AUC of 0.87 on the external validation cohort.
- This model accurately predicted 2-year FCHD risk with an AUC of 0.91 and showed high sensitivity (69%) and specificity (89%).
Conclusions:
- High-accuracy prediction of 2-year FCHD risk is achievable using single-lead ECG data.
- Combining ECG-AI predictions, particularly from Lead I, with demographic information significantly enhances FCHD risk prediction accuracy.
- These findings support the potential of accessible ECG technology for proactive FCHD risk assessment.
Abstract:
Background: Fatal coronary heart disease (FCHD) affects ~650,000 people yearly in the US. Electrocardiographic artificial intelligence (ECG-AI) models can predict adverse coronary events, yet their application to FCHD is understudied. Objectives: The study aimed to develop ECG-AI models predicting FCHD risk from ECGs. Methods (Retrospective): Data from 10 s 12-lead ECGs and demographic/clinical data from University of Tennessee Health Science Center (UTHSC) were used for model development. Of this dataset, 80% was used for training and 20% as holdout. Data from Atrium Health Wake Forest Baptist (AHWFB) were used for external validation. We developed two separate convolutional neural network models using 12-lead and Lead I ECGs as inputs, and time-dependent Cox proportional hazard models using demographic/clinical data with ECG-AI outputs. Correlation of the predictions from the 12- and 1-lead ECG-AI models was assessed. Results: The UTHSC cohort included data from 50,132 patients with a mean age (SD) of 62.50 (14.80) years, of whom 53.4% were males and 48.5% African American. The AHWFB cohort included data from 2305 patients with a mean age (SD) of 63.04 (16.89) years, of whom 51.0% were males and 18.8% African American. The 12-lead and Lead I ECG-AI models resulted in validation AUCs of 0.84 and 0.85, respectively. The best overall model was the Cox model using simple demographics with Lead I ECG-AI output (D1-ECG-AI-Cox), with the following results: AUC = 0.87 (0.85-0.89), accuracy = 83%, sensitivity = 69%, specificity = 89%, negative predicted value (NPV) = 92% and positive predicted value (PPV) = 55% on the AHWFB validation cohort. For this, the 2-year FCHD risk prediction accuracy was AUC = 0.91 (0.90-0.92). The 12-lead versus Lead I ECG FCHD risk prediction showed strong correlation (R = 0.74). Conclusions: The 2-year FCHD risk can be predicted with high accuracy from single-lead ECGs, further improving when combined with demographic information.
