Key result
An AI ensemble model detects obstructive CAD on non-contrast CT with ~76% accuracy.
Why the study?
Early identification of obstructive coronary artery disease is crucial because it is strongly associated with acute myocardial infarction, prompting development of an ensemble model combining deep learning and machine learning.
Does a deep learning and machine learning ensemble model improve the detection of obstructive coronary artery disease compared to individual models in patients undergoing cardiac CT?
Population
1054 patients
Comparison
Deep learning and machine learning ensemble model vs individual component models
Design
Retrospective cohort feasibility study
Authors
Loading...
May aid ObCAD detection in retrospective cohorts; leaves open prospective validation and clinical impact.
Cohort (n=1,054)
Does a deep learning and machine learning ensemble model improve the detection of obstructive coronary artery disease compared to individual models in patients undergoing cardiac CT?
Effect estimate: ROC AUC 0.81 ± 0.04
An ensemble model combining deep learning on non-contrast CT scans with machine learning on clinical features enhances the detection of obstructive coronary artery disease.
Park et al. (2026) conducted a cohort in Obstructive coronary artery disease (ObCAD) (n=1,054). Ensemble model integrating deep learning and machine learning vs. Individual component models was evaluated on Binary classification of ObCAD (>50% stenosis) (ROC AUC 0.81 ± 0.04). An ensemble model integrating deep learning and machine learning detected obstructive coronary artery disease with a mean ROC AUC of 0.81 ± 0.04 and an accuracy of 0.76 ± 0.04.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: