Construction of a Personalized Prediction Model for Diabetic Lung Infection Based on Machine Learning
Authors' conclusion
We developed a robust and clinically applicable nomogram for predicting the risk of pneumonia in diabetic patients with pulmonary infections. This model exhibits high accuracy and may assist clinicians in identifying high-risk individuals who could benefit from early preventive measures and timely interventions.
Study context
- Population
- Patients diagnosed with pulmonary infections at one hospital, both diabetic and nondiabetic
- Participants
- 168 patients
- Comparison
- Training vs Validation
- Outcome
- Risk of lung infection in diabetic patients (nomogram discrimination, calibration, clinical utility)
Methodological detailssingle-centre, least absolute shrinkage and selection operator (LASSO) feature selection, multivariate logistic regression nomogram, internal validation
- Single-centre
- LASSO feature selection
- Multivariate logistic regression nomogram
- Internal validation
Performance
As reported in the abstract
Nomogram discrimination (AUC)
| Test or model | AUC |
|---|---|
| Training | 0.919 (95%CI: 0.825-0.937) |
| Validation | 0.862 (95% CI: 0.819-0.912) |
AUC
Training0.919 (95%CI: 0.825-0.937)
Validation0.862 (95% CI: 0.819-0.912)
Original publication
Construction of a Personalized Prediction Model for Diabetic Lung Infection Based on Machine Learning. Clin Endocrinol (Oxf). Published 14 September 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →