Explainable coronary artery disease prediction model based on AutoGluon from AutoML framework.

Area of Science:

• Cardiovascular Medicine
• Artificial Intelligence
• Machine Learning

Background:

• Clinical diagnosis of Coronary Artery Disease (CAD) requires accurate predictive models.
• Traditional model development can be complex and time-consuming.
• Need for explainable AI in medical decision-making.

Purpose of the Study:

• To apply Automated Machine Learning (AutoML) for an explainable CAD prediction model.
• To support clinical diagnosis in cardiovascular medicine.
• To evaluate the feasibility and performance of AutoML in this domain.

Main Methods:

• Utilized a combined dataset from five public CAD-related sources.
• Developed an ensemble model using the AutoGluon AutoML framework.
• Explained model predictions using SHapley Additive exPlanations (SHAP).

Main Results:

• The AutoGluon ensemble model outperformed individual baseline models in CAD prediction.
• Achieved high performance metrics: 0.9167 accuracy and 0.9562 AUC (4-fold cross-bagging).
• SHAP analysis provided feature importance and explained model predictions.

Conclusions:

• AutoML is feasible and effective for cardiovascular disease prediction.
• AutoML simplifies model building and enhances prediction accuracy.
• SHAP integration improves model transparency and credibility for clinical adoption.