Detection of Obstructive Coronary Artery Disease Using a Deep Learning and Machine Learning Ensemble: A Retrospective Feasibility Study
Doyoung Park, Linxuan Yan, Arman Ahmad Khan, Wilbert Hsien Hao Ho, Choon Ta Ng, Jonathan Jiunn Liang Yap, Swee Yaw Tan, Khung Keong Yeo, Lohendran BaskaranAbstract
Aims
Early identification of obstructive coronary artery disease (ObCAD) is crucial because it is strongly associated with acute myocardial infarction. We developed a weighted average ensemble model integrating deep learning (DL) and machine learning (ML) to leverage imaging and clinical data for enhancing the detection of ObCAD.
Methods and results
A retrospective cohort of 1,054 patients was used to develop an ensemble model combining a 3D Vision Transformer (ViT) with eXtreme Gradient Boosting and CatBoost for binary classification of ObCAD (>50% stenosis). Unstructured data comprised 3D cardiac non-contrast computed tomography scans, while structured data included 11 demographic and clinical features. ObCAD labels were derived from corresponding coronary computed tomography angiography. Model performance was evaluated using 10-fold cross-validation with fold-wise Wilcoxon signed-rank testing. The ensemble model achieved a mean area under the receiver operating characteristic curve (ROC AUC) of 0.81 ± 0.04 and an accuracy of 0.76 ± 0.04. It demonstrated a statistically significantly higher ROC AUC than individual component models. Feature importance analysis identified age, chest pain, and sex as the most influential predictors of ObCAD. Grad-CAM visualization indicated that the 3D ViT primarily focused on cardiac regions containing coronary artery calcium deposits.
Conclusion
Integrating DL-based imaging analysis with ML-based clinical modeling enhances the discriminative performance for ObCAD detection with complementary interpretability. This ensemble framework demonstrates potential to support clinical decision-making by identifying high-risk patients using routine cardiac CT combined with patient-level clinical data. Future studies using external validation and coronary artery calcium scores may further improve risk prediction.