Predicting major amputation risk in diabetic foot ulcers using comparative machine learning models for enhanced...

Area of Science:

• Medical Informatics
• Machine Learning in Healthcare
• Diabetology

Background:

• Diabetic foot ulcers (DFU) pose a significant risk for major amputation.
• Prompt identification of high-risk DFU patients is crucial for timely intervention.
• Existing predictive methods may lack accuracy or efficiency.

Purpose of the Study:

• To develop and validate a machine learning model for early prediction of major amputation risk in DFU patients upon admission.
• To identify key clinical and laboratory features associated with major amputation in DFU.

Main Methods:

• Utilized synthetic minority oversampling technique for class imbalance.
• Developed six machine learning models (logistic regression, random forest, SVM, KNN, GBM, XGBoost) using 17 identified features.
• Evaluated model performance using accuracy, precision, recall, F1-score, and AUC.

Main Results:

• The Gradient Boosting Machine (GBM) model demonstrated superior performance (Accuracy: 0.9408, AUC: 0.9499).
• Key predictors for major amputation included multidrug-resistant infection, C-reactive protein (CRP), diabetes duration, troponin (Tn), and age.
• The GBM model effectively predicted the risk of major amputation in DFU patients.

Conclusions:

• Gradient Boosting Machine is an effective tool for predicting major amputation risk in diabetic foot ulcer patients.
• Early identification via machine learning can guide clinical decision-making and potentially reduce amputation rates.
• Highlighting specific risk factors aids in targeted patient management strategies.