[Retracted] Predicting the Spread of Vessels in Initial Stage Cervical Cancer through Radiomics Strategy Based on Deep Learning Approach

Table 2

Various CNN model performance evaluation [A_c = Accuracy; S_n = Sensitivity; S_p = Specificity].

Various methods	ROC curve	Ac	Sn	Sp
AdaptiveVGG19 X1DCE	0.79	0.75	0.65	0.78
AdaptiveVGG16 X1DCE	0.82	0.68	0.74	0.75
AdaptiveInceptionV3/X1DCE	0.78	0.81	0.85	0.65
AdaptiveVGG16/X2WI	0.85	0.68	0.88	0.60
AdaptiveResNet50-V2/X1DCE	0.89	0.78	0.68	0.87
AdaptiveInceptionV3/X2WI	0.92	0.69	0.66	0.88
AdaptiveResNet50-V2/X2WI	0.88	0.80	0.70	0.92
AdaptiveDenseNet121/X2WI	0.87	0.88	0.72	0.77
AdaptiveDenseNet121/X1DCE	0.92	0.77	0.87	0.84
AdaptiveVGG19-SE/X2WI	0.89	0.85	0.79	0.82
AdaptiveVGG19-SE/X1DCE	0.94	0.89	0.88	0.88
AdaptiveVGG19-CBAM/X2WI	0.85	0.91	0.82	0.89
AdaptiveVGG19-CBAM/X1DCE	0.90	0.87	0.79	0.88
Proposed A-EL/X1DCE + X2WI	0.95	0.96	0.91	0.94