Table 4 Delong’s test results of major predictors’ AUCs for NE vs. SKC.

From: Detection of subclinical keratoconus using a novel combined tomographic and biomechanical model based on an automated decision tree

 

AUC difference

SE

95% CI

P value

BAD-D vs. CBI

0.06

0.04

− 0.02–0.15

0.158

BAD-D vs. TBI

0.02

0.03

− 0.04–0.09

0.420

TBI vs. CBI

0.04

0.05

− 0.06–0.13

0.422

CHAID vs. BAD-D

0.07

0.03

0.01–0.13

0.014

CHAID vs. CBI

0.14

0.04

0.05–0.22

0.001

CHAID vs. TBI

0.10

0.03

0.03–0.16

0.004

CART vs. BAD-D

0.08

0.03

0.02–0.14

0.015

CART vs. CBI

0.14

0.05

0.05–0.23

0.002

CART vs. TBI

0.10

0.03

0.04–0.17

0.002

CHAID vs. CART

0.01

0.02

− 0.03–0.05

0.781

  1. BAD-D Belin/Ambrósio Enhanced Ectasia Display Index, CBI corvis biomechanical index, TBI combined tomography and biomechanical index, CHAID CHAID model based on training database, CHAID CART model based on training database.
  2. Significant values are in bold.
Back to article page