Table 5. Training results when models run.

1 vs 1	Models	Accuracy	Precision	Recall	Auc	F1 score	Validation accuracy
CN/AD	EfficientNetB0	0.9990	0.9985	0.9985	0.9999	0.9984	0.9892
	DenseNet121	0.9959	0.9939	0.9939	0.9996	0.9938	0.9850
	AlexNet	0.9980	0.9970	0.9970	0.9999	0.9970	0.9896
MCI/AD	EfficientNetB0	0.9987	0.9980	0.9980	1.0000	0.9980	0.9828
	DenseNet121	0.9972	0.9958	0.9958	0.9999	0.9957	0.9406
	AlexNet	0.9958	0.9937	0.9937	0.9989	0.9937	0.9841
CN/MCI	EfficientNetB0	0.9975	0.9963	0.9963	0.9996	0.9963	0.9724
	DenseNet121	0.9869	0.9869	0.9869	0.9987	0.9968	0.9767
	AlexNet	0.9961	0.9942	0.9942	0.9986	0.9942	0.9702