Table 3.

Evaluation metrics for all tested multi-class classifiers on microscopic blood images.

Classification Model	Class	Precision	Recall (Sensitivity)	F1-Score	Accuracy
VGG-16	ALL	0.86	0.83	0.85	0.8430
	AML	0.85	0.73	0.79
	Normal	0.83	0.89	0.86
ResNet-50	ALL	0.89	0.92	0.90	0.9101
	AML	1.00	0.80	0.89
	Normal	0.90	0.95	0.92
DenseNet-121	ALL	0.87	0.94	0.91	0.9213
	AML	1.00	0.87	0.93
	Normal	0.95	0.92	0.93
Developed GAN Classifier	ALL	0.90	1.00	0.95	0.9550
	AML	1.00	0.87	0.93
	Normal	1.00	0.95	0.97