A Machine-Learning-Based Clinical Decision Model for Predicting Amputation Risk in Patients with Diabetic Foot...

Area of Science:

• Medical Informatics
• Diabetology
• Vascular Surgery

Background:

• Diabetic foot ulcers (DFUs) are a major cause of lower limb amputation.
• Predicting amputation risk is crucial for timely intervention and prevention strategies.

Purpose of the Study:

• To develop a reliable machine learning (ML) model for predicting lower limb amputation risk in DFU patients.
• To provide quantitative evidence for clinical decision-making and personalized prevention strategies.

Main Methods:

• Retrospective analysis of 149 hospitalized DFU patients.
• Feature selection using the least absolute shrinkage and selection operator (LASSO) algorithm.
• Support vector machine (SVM) model trained and validated with five-fold cross-validation.

Main Results:

• C-reactive protein and Wagner grade identified as independent predictors of amputation (p < 0.05).
• Optimized SVM model achieved an area under the ROC curve of 0.89.
• Model demonstrated good predictive performance with 82.4% correct classification in internal validation.

Conclusions:

• C-reactive protein and Wagner grade are key determinants of amputation risk in DFU patients.
• The developed SVM model offers strong accuracy and clinical interpretability for predicting amputation risk.
• The model supports personalized interventions and improved clinical decision-making for DFU management.