Autoinflammatory & rare systemic diseases
Development and assessment of an assisted diagnosis model using machine learning for identifying adult-onset Still’s disease in fever of unknown origin: a retrospective study in China
Authors' conclusion
We developed a machine learning model using electronic medical records to identify AOSD among patients with FUO with clinical suspicion for AOSD.
Study context
- Population
- Patients with fever of unknown origin (FUO) with clinical suspicion for Adult-onset Still's disease (AOSD), China, 2010–2020771 derivation, 75 external validation
- Participants
- 847 patients
- Comparison
- Logistic regression (LR) vs random forest vs extreme gradient boosting models
- Outcome
- Model performance (AUC, ROC curves, sensitivity, specificity, accuracy) for identifying AOSD
Methodological detailsretrospective, machine learning, stratified random sampling into training and internal test set (7:3), fivefold cross-validation, external validation at another centre, Six features (age, neutrophil percentage, WBC count, infection indicator, ferritin, 'AOSD-related clinical presentation score') selected by at least two of Boruta, recursive feature elimination and LASSO
- Retrospective
- Machine learning
- Stratified random sampling into training and internal test set (7:3)
- Fivefold cross-validation
- External validation at another centre
- Six features (age, neutrophil percentage, WBC count, infection indicator, ferritin, 'AOSD-related clinical presentation score') selected by at least two of Boruta, recursive feature elimination and LASSO
Performance
As reported in the abstract
AUC for identifying AOSD
| Test or model | AUC, training set (95% CI 0.956 to 0.983) | AUC, test set (95% CI 0.934 to 0.985) |
|---|---|---|
| LR | 0.969 | 0.960 |
LR
AUC, training set (95% CI 0.956 to 0.983)0.969
AUC, test set (95% CI 0.934 to 0.985)0.960
Further findings
- External validation of LR-based nomogram, AUC
- 0.906 · AUC, independent validation cohort
Original publication
Development and assessment of an assisted diagnosis model using machine learning for identifying adult-onset Still’s disease in fever of unknown origin: a retrospective study in China. RMD Open. Published 7 July 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →