Abstract
An early and accurate diagnosis of Alzheimer’s disease (AD) is crucial for implementing effective interventions, as this condition poses a significant global health risk. This study presents a novel model, Residual Alzheimer-Net+, which analyzes MRI and PET datasets to address the challenges of AD classification. By integrating data from both imaging modalities, this robust and user-friendly classification system enhances the understanding of AD. The model employs deep neural network architecture with residual connections to optimize information flow and mitigate issues related to vanishing gradients, ensuring effective learning from multi-modal data. Experimental results demonstrate that Residual Alzheimer-Net+ can identify complex patterns indicative of AD across various imaging datasets, achieving superior training, testing, and validation accuracy compared to existing methods.
Keywords: AD, MRI, Multi-Modal Classification, PET, Residual Alzheimer- Net+