Table 1 Comparison with existing methods using ADNI database for AD vs CN classification.

Gonuguntla et al.³⁵

65AD + 65CN

ROIs

85.3

–

–

Kang et al.³⁶

187AD + 229CN

DCGAN

84.23

–

–

Vgg16

83.57

–

–

ResNet50

74.7

–

–

Kushol et al.³⁷

159AD + 229CN

Transformer

88.2

95.6%

77.4%

Cai et al.³⁸

49AD + 43CN

Graph Transformer

85.9

79.8%

92.8%

Qiu et al.²³

188AD + 229CN

FCN

83.4

76.7%

88.9%

Shahamat et al.³⁹

70AD + 70CN

3D CNN

85

–

–

Shojaei et al.⁴⁰

74AD + 71CN

3D CNN

87

–

–

Tinauer et al.⁴¹

128AD + 290CN

3D CNN

86.19

79.73%

92.66%

Proposed method

188AD + 229CN

3D CNN

90.91

80.95%

100%