Skip to main content
. 2018 Sep 5;12:596. doi: 10.3389/fnins.2018.00596

Table 3.

The result of ROC analysis and Logistic regression analysis.

AUC Confidence interval p Cut-off value
L 0.673 0.582–0.763 0.001 13.5
H 0.750 0.673–0.828 <0.001 12.4
AR 0.847 0.760–0.897 <0.001 1.6
SR 0.824 0.754–0.893 <0.001 2.3
EI 0.768 0.692–0.843 <0.001 0.11
UI 0.691 0.608–0.775 <0.001 0.13
NSI 0.877 0.820–0.935 <0.001 0.17
DA 0.862 0.807–0.941 <0.001 35°
NWSSa 0.859 0.779–0.910 <0.001 0.24
NWSSm 0.784 0.720–0.869 <0.001 0.38
WSSG 0.702 0.618–0.796 0.004 15.0
LSAR 0.749 0.584–0.773 <0.001 0.30
OSI 0.843 0.660–0.789 <0.001 0.008
RRT 0.654 0.499–0.761 <0.001 5.3

Variable OR Univariate logistic regression p value OR Multivariate logistic regression p value

L 1.40 0.344
AR 1.63 0.011 11.72 0.053
SR 12.59 <0.001 3.46 0.001
H 1.70 0.172
DA 5.86 <0.001 4.42 <0.001
EI 3.92 0.028 12.76 0.071
UI 19.84 0.056
NSI 7.90 <0.001 3.17 0.072
WSSG 2.04 0.426
NWSSa 27.76 <0.001 0.58 0.032
NWSSm 0.087 0.027 0.03 0.487
LSAR 2.87 0.003 3.65 0.007
OSI 1.85 0.007 0.895 0.032
RRT 4.32 0.059

Bold values indicate statistically significant parameters.