A Convolutional Vision Transformer with Channel Attention for Multi-Class Alzheimer's Disease Classification Using MRI

Alzheimer's disease is a progressive neurodegenerative disorder and the primary cause of dementia globally. Early and accurate diagnosis is essential for improving patient outcomes. This paper proposes a novel deep learning architecture that integrates a Convolutional Vision Transformer with Squeeze-And-Excitation blocks to classify different stages of Alzheimer's disease from MRI images. Unlike prior studies, which often merge moderate and mild dementia classes, this work evaluates the model in binary, three-class (excluding the minority class), and full four-class settings. Experimental results demonstrate that the proposed model achieves an accuracy of 99.36 %, outperforming or matching state-of-The-Art approaches even without data balancing. These results underscore the potential of combining attention mechanisms with transformer-based architectures for enhancing automated AD diagnosis and provide a promising step toward more clinically robust solutions.