Development and temporal validation of an explainable machine learning model for predicting in-hospital diabetic foot ulcer in patients with type 2 diabetes using inflammatory–nutritional composite in
Source check: Source checked. Not independent scientific confirmation.
Finding: The Frontiers in Medicine study develops explainable ML models for in-hospital diabetic foot ulcer in a 550-patient cohort with a temporal validation split, a clinically specific retrospective assessment, though the validation cohort has only 18 events.
Technology: Logistic regression, random forest, LightGBM and XGBoost models used routinely available admission variables and inflammatory–nutritional indices to estimate in-hospital diabetic-foot-ulcer risk; LASSO selected predictors and SHAP interpreted the models.
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Logistic regression, random forest, LightGBM and XGBoost models used routinely available admission variables and inflammatory–nutritional indices to estimate in-hospital diabetic-foot-ulcer risk; LASSO selected predictors and SHAP interpreted the models.
- Test and result
- In 550 hospitalized patients with type 2 diabetes (400 development, 150 temporal validation), logistic regression had temporal-validation AUC 0.976, but calibration was poor (Hosmer–Lemeshow P=.001); LightGBM/XGBoost sensitivities were 0.056/0.389 at prespecified thresholds.
- Limitation / still unknown
- The temporal validation set contained 18 events; low sensitivity and calibration concerns preclude screening use. Multicenter prospective validation is still needed.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2026.1938429/full
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.