Efficiency of Disease and Disease Activity Diagnosis Models of Systemic Lupus Erythematosus Based on Protein Array Analysis

Figure 2

Efficiency and validation of the disease diagnosis model. (a) ROC of diagnosis model in differentiating SLE patients from healthy controls. (b) Calibration curve of the diagnosis model. (c) Nomogram based on the diagnosis model. (d) Decision curve for the diagnosis model in predicting SLE. Standard net benefit (-axis) and risk threshold (-axis) formed the coordinate system. The red line represented our model, grey line represented the assumption that all the people were suffering SLE, and black line represented the assumption that all the people were healthy. (e) Clinical impact curve analysis diagram. Red line represented number of people which were diagnosed with SLE by our model at different threshold probability, and blue line represented number of SLE patients.

(a)

(b)

(c)

(d)

(e)