Table 2.

Classification results with data augmentation.

CNN model	Accuracy (%)	Sensitivity (%)	Specificity (%)	$F_1$ score	AUC
AlexNet	74.50 ± 4.40	70.46 ± 6.37	79.05 ± 8.61	0.75 ± 0.04	0.83 ± 0.04
GoogLeNet	78.97 ± 3.70	75.95 ± 13.69	82.38 ± 10.53	0.79 ± 0.06	0.91 ± 0.04
SqueezeNet	78.52 ± 7.56	91.56 ± 7.63	63.81 ± 23.79	0.82 ± 0.04	0.90 ± 0.01
ShuffleNet	86.13 ± 10.16	83.54 ± 19.89	89.05 ± 5.77	0.86 ± 0.12	0.93 ± 0.06
ResNet-18	90.16 ± 2.36	89.45 ± 7.31	90.95 ± 9.29	0.91 ± 0.02	0.96 ± 0.05
ResNet-50	92.62 ± 4.19	91.14 ± 3.35	94.29 ± 5.15	0.93 ± 0.04	0.98 ± 0.01
ResNet-101	89.71 ± 10.05	82.28 ± 20.09	98.10 ± 2.18	0.89 ± 0.12	0.97 ± 0.03
Xception	85.68 ± 6.76	90.72 ± 4.79	80.00 ± 19.64	0.87 ± 0.05	0.94 ± 0.04
Inception-v3	91.28 ± 8.25	90.30 ± 5.12	92.38 ± 11.98	0.92 ± 0.08	0.97 ± 0.02
Inception-ResNet-v2	86.35 ± 5.71	88.19 ± 6.37	84.29 ± 14.50	0.87 ± 0.05	0.95 ± 0.05
VGG-16	78.52 ± 10.02	74.68 ± 30.14	82.86 ± 15.91	0.76 ± 0.17	0.91 ± 0.04
VGG-19	83.22 ± 5.85	90.72 ± 3.19	74.76 ± 12.96	0.85 ± 0.04	0.90 ± 0.05
DenseNet-201	91.72 ± 6.52	88.61 ± 8.86	95.24 ± 4.36	0.92 ± 0.07	0.97 ± 0.03
MobileNet-v2	87.25 ± 10.46	95.78 ± 2.64	77.62 ± 21.63	0.89 ± 0.08	0.95 ± 0.04
NasNet-Mobile	83.45 ± 7.36	84.81 ± 2.19	81.90 ± 17.46	0.85 ± 0.05	0.94 ± 0.04
NasNet-Large	85.23 ± 8.25	79.32 ± 16.28	91.90 ± 5.77	0.84 ± 0.10	0.93 ± 0.05