Table 8.

Quantitative performance analysis of the proposed architecture with state-of-the-art methods for AD classification.

S. No	Model Trained	Number of Trainable Parameters	Accuracy in (%)
1	AlexNet	57 M	64.69
2	EfficientNet-B0	4 M	75.01
3	ResNet-18	11 M	78.69
4	VGGNet-16	134 M	78.82
5	DarkNet	26 M	80.43
6	DenseNet-169	12 M	82.2
7	AHANet (Proposed)	32 M	98.53