Figure 4.
The validation of neural network NN01 in patients with APS (n = 33), controls (n = 62), and thrombosis controls (n = 38). (A-B) Confusion matrices showing the accuracy of diagnosis in patients with APS compared with an anticoagulated control cohort (A) and a thrombosis control cohort (B). (C-D) ROC curves of the diagnostic accuracy of NN01 to identify patients with APS from an anticoagulated control cohort (C) and a thrombosis control cohort (D). (E-F) Quantification of APS suspicion by NN01 according to subject type in anticoagulated controls (E) and thrombosis controls (F). (G) Quantified sensitivity and specificity of NN01 for the diagnosis of APS. (H) Positive and negative predictive values of NN01.

The validation of neural network NN01 in patients with APS (n = 33), controls (n = 62), and thrombosis controls (n = 38). (A-B) Confusion matrices showing the accuracy of diagnosis in patients with APS compared with an anticoagulated control cohort (A) and a thrombosis control cohort (B). (C-D) ROC curves of the diagnostic accuracy of NN01 to identify patients with APS from an anticoagulated control cohort (C) and a thrombosis control cohort (D). (E-F) Quantification of APS suspicion by NN01 according to subject type in anticoagulated controls (E) and thrombosis controls (F). (G) Quantified sensitivity and specificity of NN01 for the diagnosis of APS. (H) Positive and negative predictive values of NN01.

Close Modal

or Create an Account

Close Modal
Close Modal