Machine Learning Approach for Predicting Systemic Lupus Erythematosus in Oman-based Cohort

Abstract

Objectives: Design a machine learning-based prediction framework to predict the presence or absence of Systemic Lupus Erythematosus (SLE) in a cohort of Omani patients. Methods: Records of 219 patients from 2006 to 2019 were extracted from SQU Hospital electronic records, 138 patients have SLE, and the remaining 81 have other rheumatologic diseases. Clinical and demographic features were analyzed to focus on the early stages of the disease. Our design implements Recursive Feature Selection (RFE) to select only the most informative features. In addition, the CatBoost classification algorithm is utilized to predict SLE and an explainer algorithm (SHAP) is applied on top of the CatBoost model to provide individual prediction reasoning which is then validated by rheumatologists. Results: CatBoost achieved an Area Under the ROC curve (AUC) score of 0.95 and a Sensitivity of 92%. Four clinical features (Alopecia, renal disorders, Acute Cutaneous Lupus, and hemolytic anemia) along with the patient’s age were shown to have the greatest contribution to the prediction by the SHAP algorithm. Conclusion: We have designed and validated an explainable framework to predict SLE patients and provide reasoning for its prediction. Our framework enables early intervention for clinicians which leads to positive healthcare outcomes.

Keywords: Systemic Lupus Erythematosus; Interpretation; Machine Learning; Supervised; Clinical Decision Support System; Statistical Data; Data Analysis.