Risk Prediction of Diabetic Foot Amputation Using Machine Learning and Explainable Artificial Intelligence
Authors' conclusion
Machine learning algorithms performed well in predicting the risk of LEA in a patient with DFU. Explainability can help provide clinical insights and identify at-risk patients for early intervention.
Study context
- Population
- Inpatient DFU episodes in a tertiary institution, 2012–2017
- Participants
- 2,559 inpatient DFU episodes
- Outcome
- Risk of major LEA, minor LEA and any LEA (model performance by ROC, balanced-accuracy, F1-score)
Methodological detailsmachine learning prediction model, explainable AI (SHAP), single tertiary institution, 51 features reviewed: demographics, comorbidities, medication, wound characteristics, laboratory results
- Machine learning prediction model
- Explainable AI (SHAP)
- Single tertiary institution
- 51 features reviewed: demographics, comorbidities, medication, wound characteristics, laboratory results
Main finding
As reported in the abstract
Model performance (ROC) for LEA prediction
ROC for major LEA (best: XGBoost)
0.820
ROC for minor LEA (best: XGBoost)
0.637
ROC for any LEA (best: Gradient Boosted Trees)
0.756
Observational data: associations, not effects.
Further findings
Key features by SHAP
- Total white cell count, comorbidity score and red blood cell count contributed highest weightage to major LEA
- Total white cell, eosinophils and necrotic eschar in the wound contributed most to any LEA
Original publication
Risk Prediction of Diabetic Foot Amputation Using Machine Learning and Explainable Artificial Intelligence. J Diabetes Sci Technol. Published 30 January 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →