Identification of key factors and explainability analysis for surgical decision-making in hepatic alveolar echinococcosis assisted by machine learning

Figure 1 Feature selection process. A: Recursive feature elimination (RFE) algorithm results; B: Least absolute shrinkage and selection operator (LASSO) regression feature selection; C: Intersection of features selected by RFE, minimum redundancy maximum relevance, and LASSO. RFE: Recursive feature elimination; LASSO: Least absolute shrinkage and selection operator; mRMR: Minimum redundancy maximum relevance.

Figure 2 Model performance and validation. A: Receiver operating characteristic (ROC) curves of 11 machine learning models; B: ROC curves of the eXtreme Gradient Boosting (XGBoost) model on the training and validation sets after parameter optimization; C: Calibration curve of the XGBoost model, showing consistency between predicted probabilities and observed proportions; D: Decision Curve Analysis of the XGBoost model, demonstrating net clinical benefit. ROC: Receiver operating characteristic; XGBoost: EXtreme Gradient Boosting; DCA: Decision curve analysis; kNN: K-Nearest Neighbors; SVM: Support Vector Machine; GP: Gaussian Process; LR: Logistic Regression; MN: Neural Network; RF: Random Forest; GBM: Gradient Boosting Machine; C5.0: C5.0 Decision Tree; Ada: AdaBoost.

Figure 3 SHapley Additive exPlanations analysis for model interpretability. A: SHapley Additive exPlanations (SHAP) summary bar chart, displaying the average SHAP values of features in descending order; B: SHAP summary dot plot, showing the direction and magnitude of each feature's influence on the model's prediction; C: SHAP waterfall chart for the 10^th hepatic alveolar echinococcosis patient, illustrating the contribution of each feature to the prediction result. SHAP: SHapley Additive exPlanations; XGBoost: EXtreme Gradient Boosting; WBC: White blood cell count; PLT: Platelet count; Hb: Hemoglobin; IBil: Indirect bilirubin; GGT: γ-glutamyl transferase; Scr: Serum creatinine; PT: Prothrombin time.