Table 2.

Comparison of proposed framework with the existing CNN models.

Models	Accuracy %	F1-score	Precision	MCC	Recall	AUC-PR	AUC-ROC
AlexNet	92.86	0.6807	0.9960	0.5874	0.5171	0.9041	0.9685
VGG16	94.72	0.9146	0.9552	0.839	0.8772	0.9321	0.9816
Inceptionv3	94.89	0.8055	0.9920	0.7091	0.6780	0.8972	0.9860
VGG19	95.38	0.8353	0.9902	0.7429	0.7223	0.9088	0.9739
Resnet50	95.62	0.8282	0.9971	0.7379	0.7082	0.9432	0.9848
Shufflenet	95.93	0.8491	0.9949	0.7621	0.7404	0.9541	0.9901
DenseNet201	96.17	0.8685	0.9917	0.7856	0.7726	0.9471	0.9884
Xception	96.57	0.9342	0.9737	0.8651	0.9074	0.9527	0.9882
GoolgeNet	96.72	0.8934	0.9917	0.8195	0.8128	0.9469	0.9881
Proposed iMDA	97.93	0.9394	0.9864	0.8796	0.8873	0.9731	0.9938