Application of machine learning in assessing disease activity in SLE
Yun Wang1, Peihong Yuan1, Wei Wei1
1Department of Laboratory Medicine, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China.
Lupus Science & Medicine
|April 9, 2025
Summary
A new machine learning model uses objective lab tests to accurately assess Systemic Lupus Erythematosus (SLE) disease activity. This approach offers a more feasible and timely evaluation than subjective methods.
Area of Science:
- Immunology
- Computational Biology
- Medical Informatics
Background:
- Systemic Lupus Erythematosus (SLE) is a chronic autoimmune disease characterized by immune complex deposition and inflammation.
- Current disease activity assessment, like the Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K), relies on subjective clinical judgment.
- There is a need for objective and reliable methods to evaluate SLE disease activity.
Purpose of the Study:
- To develop and validate a machine learning (ML) model for assessing SLE disease activity using objective laboratory indicators.
- To improve the objectivity and feasibility of SLE disease activity evaluation in clinical practice.
Main Methods:
- A retrospective study involving 319 SLE patients was conducted.
- Clinical characteristics and laboratory indicators were collected for model development.
- Multiple ML algorithms were applied, with XGBoost selected for its superior performance.
Main Results:
- Six key laboratory indicators were identified: anti-dsDNA (IFT), quantitative anti-dsDNA, neutrophils, globulin, proteinuria, and NK cells.
- The XGBoost model achieved high performance in distinguishing active SLE.
- Performance metrics included an AUC of 0.934, accuracy of 0.925, sensitivity of 0.969, specificity of 0.750, and F1 score of 0.954.
Conclusions:
- A novel ML model utilizing objective laboratory data provides a feasible method for assessing SLE disease activity.
- This model can potentially enable timely evaluation, aiding treatment decisions and prognosis.
- This represents a significant advancement over subjective assessment tools.
