A Differential Diagnostic Model for Moyamoya Disease and Non- moyamoya Ischemic Stroke: A Highly Efficient Clinical Approach

Abstract

Abstract Purpose: Many moyamoya disease (MMD) ischemic strokes were misdiagnosed and could not be treated effectively. To address this question, we aimed to screen easily obtainable clinical variables to construct a differential diagnostic model between MMD and non-MMD ischemic stroke. Methods: A total of 300 patients (150 MMD and 150 non-MMD ischemic strokes) in Henan Provincial People's Hospital were selected and divided into training (210) and validation cohorts (90). Binary logistic regression analysis, lasso regression, and support vector machine (SVM) were used to construct the diagnostic model. The optimal model was visualized by nomograms, and the discriminant ability of the nomogram was tested in the training and validation cohorts, respectively. Results: Among the three models, binary logistic regression has the most significant C statistic (0.87 and 0.88) in the training cohort and validation cohort, respectively. The variables that showed a significant difference in the multivariate logistic regression analysis were systolic blood pressure (SBP), total cholesterol (TC), albumin (ALB), free triiodothyronine (FT3), homocysteine (HCY), and age. The Hosmer-Lemeshow test P values of nomograms in the training and validation cohorts were 0.28 and 0.19, respectively, and the calibration curves were well corrected. Patients with nomogram scores below or above 168 were considered to have a low or high risk of ischemic stroke in MMD, respectively. Conclusion:Using nomograms to identify MMD and non-MMD ischemic stroke, the model has been validated to have a good discriminatory ability in both the training and validation cohorts, improving clinicians’ awareness of MMD ischemic stroke.