Fig. 8 | Scientific Reports

Fig. 8

From: Development and validation of a machine learning model for critical progression risk in pediatric severe community-acquired pneumonia

Fig. 8

SHAP values of XGBoost model for one patient. The SHAP force plot illustrates an individual-level prediction. Red bars represent risk-enhancing contributions, and blue bars represent protective effects. For this patient (predicted probability = 0.91), LDH = 350 U/L (SHAP + 0.38) and BUN = 3.46 mg/dL (SHAP + 0.22) increased risk, while PCT = 0.53 ng/mL (SHAP − 0.15) and RDW-CV = 14.8% (SHAP − 0.09) mitigated risk. This example demonstrates how the model integrates multiple biomarkers to provide a transparent, individualized explanation.

Back to article page