Table 6 Comparison with other methods reported in the literature.
Reference | Data | Method(s) | Accuracy | Classification Task |
|---|---|---|---|---|
Ahsan et al.20 | PET | 3D-CNN Box Filtering augmentation | 86.22% | AD-NC |
Ahsan et al.21 | PET | 3D-CNN (Scenario-2) | 86.22% | AD-NC |
Our approach | PET | 3D-CNN LoG augmentation | 87.24% | AD-NC |
Ahsan et al.21 | PET | 3D-CNN (Scenario-2) | 69.1% | AD-MCI |
Our approach | PET | 3D-CNN Prewitt-edge emphasizing augmentation | 70.68% | AD-MCI |
Ahsan et al.22 | PET | 3D-CNN | 62.25% | MCI-NC |
Ahsan et al.20 | PET | 3D-CNN Median Filtering augmentation | 64.82% | MCI-NC |
Our approach | PET | 3D-CNN Local Laplacian augmentation | 66.33% | MCI-NC |
Ahsan et al.21 | PET | 3D-CNN (Scenario-2) | 56.31% | AD-MCI-NC |
Ahsan et al.20 | PET | 3D-CNN Gaussian filtering augmentation | 55.63% | AD-MCI-NC |
Our approach | PET | 3D-CNN Prewitt-edge emphasizing augmentation | 57.68% | AD-MCI-NC |
