Fig. 4
From: The relationship between serum astroglial and neuronal markers and AQP4 and MOG autoantibodies
Multivariable logistic regression was used to separate the cases with αAQP4+ or αMOG+ from the cases with a double negative status. (A) The scatterplot of the log-transformed GFAP vs. tau. The dashed line indicates the prediction of αAQP4+ or αMOG+ versus αMOG-/αAQP4-. (B) The ROC curve has an unadjusted AUC of 0.81 (0.80 after optimism-adjustment)