Development and validation of a machine learning predictive model for one-month post-revascularization angina in patients who had undergone PCI or CABG

Figures

• 

  Figure 1. Flow diagram of the study population.

• 

  Figure 2. Predictor selection using Boruta. NYHA classification, The New York Heart Association classification; cTnT, cardiac troponin T; PT, prothrombin time; DBP, diastolic blood pressure; MPA indicates moderate physical activity; HDL, high-density lipoprotein cholesterol; FIB, fibrinogen; CRP, C-reactive protein; TT, thrombin time; APTT, activated partial thromboplastin time; LDH, lactatedehydrogenase; ApoA-I, Apolipoprotein A-I; ApoB, Apolipoprotein B; TG, triglyceride; BMI, body mass index; HCY, homocysteine; CKMB, creatine kinase-MB; TC, total cholesterol; LDL, low-density lipoprotein cholesterol; FBG, fasting blood glucose; LVEF, left ventricular ejection fraction; UA, uric acid; SBP, systolic blood pressure.

• 

  Figure 3. Receiver operating characteristic (ROC) curves for each model in the internal validation cohort. DT, Decision tree model; KNN, K-Nearest Neighbor model; LGBM, Light Gradient Boosting Machine model; LR, Logistic regression model; NB, Naive Bayes model; RF, Random forest model; XGB, Extreme gradient boosting model; SVM, Support Vector Machine.

• 

  Figure 4. Comparative performance analysis of eight models in the internal validation cohort. (A) Multi-metric performance summary of each model for PRA in the internal validation cohort. (B) Decision curve analysis (DCA) curves for each model in the internal validation cohort. PRA, post-revascularization angina.

• 

  Figure 5. Receiver operating characteristic (ROC) curves for each model in the external validation cohort. DT, Decision tree model; KNN, K-Nearest Neighbor model; LGBM, Light Gradient Boosting Machine model; LR, Logistic regression model; NB, Naive Bayes model; RF, Random forest model; XGB, Extreme gradient boosting model; SVM, Support Vector Machine.

• 

  Figure 6. Comparative performance analysis of eight models in the external validation cohort. (A) Multi-metric performance summary of each model for PRA in the internal validation cohort. (B) Calibration characteristics analysis for each model in the external validation cohort. PRA, post-revascularization angina.

• 

  Figure 7. Interpretable machine learning analysis of risk factors. (A) All samples and features are illustrated, with each row representing a feature and x-axis representing the SHAP value. The purple dots represent higher feature values, while the yellow dots represent lower feature values. (B) Ranking of variable importance based on the average value. (C) SHAP predictions for PRA samples. (D) SHAP predictions for non-PRA samples. Yellow arrows indicate a higher risk of PRA, while purple arrows indicate a lower risk of PRA. The length of the arrows helps visualize the degree of impact of the prediction, whereby the longer the arrow, the more significant the effect. SHAP, The Shapley Additive explanations; PRA, post-revascularization angina.