A Classification Tree Model with Optical Coherence Tomography Angiography Variables to Screen Early-Stage Diabetic Retinopathy in Diabetic Patients

Table 3

Validation outcomes and predictors in two models.

	Binary logistic regression model		Classification tree model
	Predicted NPDR (n)	Predicted NDR (n)	Predicted NPDR (n)	Predicted NDR (n)
True NPDR (n)	55	15	52	18
True NDR (n)	14	36	14	36
Predictors	BVCA (LogMAR) and SCP density		BVCA (LogMAR), DM duration, SCP density, and DCP density

NPDR, nonproliferative diabetic retinopathy; NDR, nondiabetic retinopathy; BCVA, best-corrected visual acuity; SCP, superficial capillary plexus; DM, diabetes mellitus; DCP, deep capillary plexus.