Lupus & antiphospholipid syndrome
Early identification of macrophage activation syndrome secondary to systemic lupus erythematosus with machine learning
Authors' conclusion
Machine learning models can significantly improve the diagnostic accuracy of MAS secondary to SLE, and the diagnostic scorecard model can facilitate personalized probabilistic predictions of disease occurrence in clinical environments.
Study context
- Population
- Patients with systemic lupus erythematosus (SLE) or macrophage activation syndrome (MAS) secondary to SLE
- Participants
- 188 patients
- Comparison
- Five machine learning models and a diagnostic scoring card for identifying MAS secondary to SLE
- Outcome
- Diagnostic performance (ROC-AUC, F1 score, F2 score) for MAS secondary to SLE
Methodological detailsretrospective, machine learning, Least Absolute Shrinkage and Selection Operator (LASSO) variable selection, external validation in four other domestic hospitals, Online assessment system established, Score results stratified into five diagnostic probability groups: extremely low (<5%), low (5-25%), normal (25-75%), high (75-95%), extremely high (>95%)
- Retrospective
- Machine learning
- LASSO variable selection
- External validation in four other domestic hospitals
- Online assessment system established
- Score results stratified into five diagnostic probability groups: extremely low (<5%), low (5-25%), normal (25-75%), high (75-95%), extremely high (>95%)
Performance
As reported in the abstract
Diagnostic performance, internal validation
| Test or model | ROC-AUC | F1 score | F2 score | Probability threshold score |
|---|---|---|---|---|
| Logistic regression (LR) | 0.998 | 0.96 | 0.952 | |
| Score card | 0.994 | 0.936 | 49 |
ROC-AUC
Logistic regression (LR)0.998
Score card0.994
F1 score
Logistic regression (LR)0.96
Score card— not stated in the abstract
F2 score
Logistic regression (LR)0.952
Score card0.936
Probability threshold score
Logistic regression (LR)— not stated in the abstract
Score card49
Further findings
| Outcome | SVM | Score card |
|---|---|---|
| Diagnostic performance, external validation | ||
| AUC | 0.947 | 0.915 |
Diagnostic performance, external validation
- Support Vector Machine (SVM) model outperformed other models in external validation
Original publication
Early identification of macrophage activation syndrome secondary to systemic lupus erythematosus with machine learning. Arthritis Res Ther. Published 9 May 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →