Table 33.

Performance results of ResNet50v2 on external dataset

	KFold	Accuracy	Precision	Recall	Specificity	F1-score
Without GAN	Fold1	0.8900	0.7738	0.8868	0.8914	0.8265
	Fold2	0.8938	0.7781	0.8958	0.8929	0.8328
	Fold3	0.8962	0.7869	0.8895	0.8990	0.8351
	Fold4	0.8858	0.7652	0.8848	0.8862	0.8207
	Fold5	0.8914	0.7735	0.8943	0.8902	0.8295
	Overall	0.8914	0.7735	0.8902	0.8919	0.8289
With GAN	Fold1	0.9189	0.8338	0.9063	0.9243	0.8685
	Fold2	0.9195	0.8302	0.9145	0.9216	0.8703
	Fold3	0.9130	0.8207	0.9028	0.9173	0.8598
	Fold4	0.9175	0.8277	0.9103	0.9206	0.8670
	Fold5	0.9147	0.8238	0.9045	0.9189	0.8623
	Overall	0.9167	0.8272	0.9077	0.9205	0.8656