Table 1.
Summary of recent Alzheimer’s classification studies.
Fig 1.
Proposed hybrid framework for Alzheimer’s disease classification.
Table 2.
Original dataset composition (Raw from ADNI).
Fig 2.
Original 2D MRI views of the brain.
Fig 3.
Gray matter segmented views of the brain.
Table 3.
GM Mask image count after preprocessing.
Table 4.
Performance metrics of the standard U-Net (Vanilla U-Net) for the brain segmentation.
Fig 4.
Visual comparison of Brain segmentation results.
Fig 5.
Proposed Multi-Layer U-Net model for gray matter segmentation.
Fig 6.
Architecture of Multi-Layer U-Net Model.
Fig 7.
Proposed Alzheimer’s disease classification pipeline using Multi-Scale EfficientNet with SVM.
Fig 8.
Architecture of Multi-EfficientNet.
Table 5.
SVM Hyperparameter Tuning Results Using Grid Search.
Table 6.
Details of Hyperparameters.
Fig 9.
XAI Saliency map visualization for the proposed model.
Table 7.
Average Saliency Map Results.
Table 8.
Comparative evaluation of gray matter segmentation using Multi-Layer U-Net, Vanilla U-Net, Attention U-Net, and state-of-the-art methods. The results of the proposed model are calculated on the test set.
Fig 10.
Gray matter segmentation results using the Attention U-Net model.
Fig 11.
Gray Matter segmentation results using the proposed Multi-layer U-Net Model.
Table 9.
Class-wise performance on the test set.
Fig 12.
Confusion Metrics: Presenting classification results on the test set.
Table 10.
Comparison of the Classification performance of the proposed model calculated on the test set with state-of-the-art Approaches.
Table 11.
Comparative performance analysis of different architectures.
Fig 13.
ROC curve for the proposed Classification model across AD, MCI, and CN.
Table 12.
Ablation Study Result.