LightGBM Predicts Coronary Risk in HHcy
Researchers at the First Affiliated Hospital of Henan University of Traditional Chinese Medicine conducted a single-center retrospective study (Jan 1, 2022–July 1, 2025) developing seven machine-learning models to predict coronary heart disease (CHD) risk in patients with hyperhomocysteinemia (HHcy). The LightGBM model performed best on the test set (AUC=0.807, F1=0.606, Brier=0.2415) using six routinely available variables. SHAP analysis highlighted age and activated partial thromboplastin time as top predictors, suggesting an interpretable tool for early risk stratification.
Key Points
- 1Develops LightGBM model achieving AUC 0.807 and F1-score 0.606 on the test set
- 2Identifies age and activated partial thromboplastin time as top predictors via SHAP analysis
- 3Enables interpretable, six-variable risk stratification usable in primary care for early CHD intervention
Scoring Rationale
Interpretable model with solid metrics and clinical relevance, limited by single-center retrospective design needing external validation.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
