Fine-tuning an ECG Foundation Model to Predict Coronary CT Angiography Outcomes

TL;DR

An interpretable AI-ECG model to predict severe or complete stenosis of the four major coronary arteries on CCTA remained stable in a clinically normal-ECG subset, indicating robustness beyond overt ECG abnormalities and Interpretability analyses revealed distinct waveform differences between high- and low-risk groups.

Abstract

Coronary artery disease (CAD) remains a major global health burden. Accurate identification of the culprit vessel and assessment of stenosis severity are essential for guiding individualized therapy. Although coronary CT angiography (CCTA) is the first-line non-invasive modality for CAD diagnosis, its dependence on high-end equipment, radiation exposure, and strict patient cooperation limits large-scale use. With advances in artificial intelligence (AI) and the widespread availability of electrocardiography (ECG), AI-ECG offers a promising alternative for CAD screening. In this study, we developed an interpretable AI-ECG model to predict severe or complete stenosis of the four major coronary arteries on CCTA. On the internal validation set, the model's AUCs for the right coronary artery (RCA), left main coronary artery (LM), left anterior descending artery (LAD), and left circumflex artery (LCX) were 0.794, 0.818, 0.744, and 0.755, respectively; on the external validation set, the AUCs reached 0.749, 0.971, 0.667, and 0.727, respectively. Performance remained stable in a clinically normal-ECG subset, indicating robustness beyond overt ECG abnormalities. Subgroup analyses across demographic and acquisition-time strata further confirmed model stability. Risk stratification based on vessel-specific incidence thresholds showed consistent separation on calibration and cumulative event curves. Interpretability analyses revealed distinct waveform differences between high- and low-risk groups, highlighting key electrophysiological regions contributing to model decisions and offering new insights into the ECG correlates of coronary stenosis.

Figures (6)

• Figure 1: ROC curves for model performance evaluation. (a) ROC curve validated using internal data from Peking University People's Hospital. (b) External validation ROC curve of the Second Hospital of Tianjin Medical University. (c) Predicted ROC curves of normal electrocardiograms from internal data.
• Figure 2: Subgroup analyses. Subgroup analysis results demonstrate that the model's predictive performance remains robust across different age groups, genders, and time intervals.
• Figure 3: Distribution of model-predicted probabilities across different degrees of coronary stenosis. Box plots illustrate the predicted probability distributions for the four major coronary arteries (LAD, LCX, LM, and RCA) across four stenosis categories (normal, mild, moderate, and severe).
• Figure 4: Calibration curve of model prediction results. The calibration curves demonstrate the model's reliability in predicting each vessel, and all four coronary arteries exhibited low Brier scores.
• Figure 5: Cumulative event occurrence curve based on model-predicted risk stratification. The cumulative event occurrence curve shows that the stratification results for each coronary artery exhibit a good separation trend and significant statistical differences.