Autoinflammatory & rare systemic diseases
Establishment of a differential diagnosis method and an online prediction platform for adult-onset Still's disease (AOSD) and sepsis based on gradient boosting decision trees algorithm
Authors' conclusion
We created a method for the identification of AOSD and sepsis based on machine learning.
Study context
- Population
- AOSD and sepsis patients admitted to Affiliated Hospital of Xuzhou Medical University, December 2018 to December 202142 AOSD, 50 sepsis
- Comparison
- Gradient boosting decision tree (GBDT) optimization model for differential diagnosis of AOSD vs sepsis
- Outcome
- Differential diagnosis of AOSD vs sepsis (AUC, accuracy, sensitivity, specificity)
Methodological detailsmachine learning model development, external validation, single-centre, Online platform AIADSS: http://cppdd.cn/STILL1/, Feature selection with SHAP; models built with scikit-learn in Python
- Machine learning model development
- External validation
- Single-centre
- Online platform AIADSS: http://cppdd.cn/STILL1/
- Feature selection with SHAP; models built with scikit-learn in Python
Performance
As reported in the abstract
Discrimination of AOSD vs sepsis by GBDT model (arthralgia, ferritin × lymphocyte count, WBC count, ferritin × platelet count, α1-acid glycoprotein/creatine kinase)
| Test or model | AUC | ACC | Sens | Spec |
|---|---|---|---|---|
| Training set | 0.9916 | 0.9457 | 0.9556 | 0.9578 |
| External validation | 0.9800 | 0.9333 | 0.8000 | 1.000 |
AUC
Training set0.9916
External validation0.9800
ACC
Training set0.9457
External validation0.9333
Sens
Training set0.9556
External validation0.8000
Spec
Training set0.9578
External validation1.000
Original publication
Establishment of a differential diagnosis method and an online prediction platform for AOSD and sepsis based on gradient boosting decision trees algorithm. Arthritis Res Ther. Published 16 November 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →