explainable prediction model for sarcopenic obesity

Explainable prediction model for sarcopenic obesity predicts sarcopenic obesity using a support vector machine model based on readily available clinical characteristics, with explainability provided through SHAP analysis.

Key Features:

• Support Vector Machine Prediction: Employs an SVM model that achieved AUC of 0.862 in internal validation and AUC of 0.785 in external validation for sarcopenic obesity prediction.
• SHAP-based Explainability: Uses SHapley Additive exPlanations (SHAP) analysis to interpret model predictions and identify the most influential predictors.
• 8-Feature Model: Predicts sarcopenic obesity using eight clinical features: BMI, gender, neck circumference, waist circumference, thigh circumference, time to full tandem standing, time to five-times sit-to-stand, and age.
• Multi-Center Validation: Developed using 1,431 participants from three community regions in Ziyang city, China, and externally validated with 832 residents from multi-center nursing homes.
• Model Comparison: Evaluated five machine learning models including support vector machine, logistic regression, random forest, light gradient boosting machine, and extreme gradient boosting.

Scientific Applications:

• Sarcopenic Obesity Screening: Enables early detection of sarcopenic obesity in aging community and nursing home populations using easily obtainable clinical measurements.
• Clinical Risk Stratification: Supports comprehensive geriatric assessments by providing interpretable predictions based on BMI, gender, and anthropometric and functional measures.
• Feature Importance Analysis: Identifies BMI and gender as the most influential predictors of sarcopenic obesity through SHAP analysis.

Methodology:

The tool uses a support vector machine model trained on eight clinical features selected after feature reduction. Model performance was assessed by the area under the receiver operating characteristic curve (AUC). SHapley Additive exPlanations (SHAP) was used for model interpretation. Sarcopenic obesity diagnosis followed the European Society for Clinical Nutrition and Metabolism (ESPEN) and the European Association for the Study of Obesity (EASO) criteria.