Table 3.

AUC Results from Dataset A of Texture Features, Shape Features, and Their Combinations under the Best Edge Detection Threshold t

AUC	Texture (T)	Shape (S)	Texture + Shape (TS)	Significance Test
NM	0.969 ± 0.002	0.971 ± 0.002 (t = 0.4)	0.976 ± 0.002 (t = 0.4)	T ≈ S, T < TS, S ≈ TS
MH	0.846 ± 0.011	0.931 ± 0.005 (t = 0.4)	0.919 ± 0.005 (t = 0.4)	T < S, T < TS, S ≈ TS
ME	0.939 ± 0.004	0.923 ± 0.005 (t = 0.3)	0.939 ± 0.004 (t = 0.4)	S < T, T ≈ TS, S < TS
AMD	0.925 ± 0.008	0.931 ± 0.005 (t = 0.2)	0.938 ± 0.006 (t = 0.2)	T ≈ S ≈ TS
Average	0.908 ± 0.008	0.932 ± 0.005	0.936 ± 0.006

Data are the mean ± SD. The best results for each macular category are shown in bold. The rightmost column shows the results of the DeLong test at a P = 0.05 significance level, where < shows that the setting on the right performs better than that on the left, and ≈ represents no significant difference.