Table 31.

Performance results of ResNet50 on external dataset

	KFold	Accuracy	Precision	Recall	Specificity	F1-score
Without GAN	Fold1	0.9003	0.7970	0.8890	0.9051	0.9404
	Fold2	0.8990	0.7911	0.8943	0.9010	0.8395
	Fold3	0.8992	0.7874	0.9025	0.8979	0.8410
	Fold4	0.8921	0.7839	0.8765	0.8987	0.8276
	Fold5	0.8981	0.7906	0.8908	0.9011	0.8377
	Overall	0.8978	0.7900	0.8906	0.9007	0.8373
With GAN	Fold1	0.9358	0.8755	0.9125	0.9456	0.8936
	Fold2	0.9317	0.8708	0.9028	0.9438	0.8865
	Fold3	0.9263	0.8539	0.9055	0.9350	0.8789
	Fold4	0.9268	0.8588	0.9003	0.9380	0.8791
	Fold5	0.9305	0.8627	0.9095	0.9393	0.8855
	Overall	0.9303	0.8644	0.9061	0.9404	0.8847