Enhancing Alzheimer Stage Classification of MRI Images through Transfer Learning
Sheshang Degadwala, Dhairya Vyas, Abhijeetsinh Jadeja, Darshanaben Dipakkumar Pandya
Abstract
This study proposes a transfer learning approach to enhance the classification of Alzheimer disease on MRI images using popular convolutional neural networks (CNNs) such as AlexNet, Vgg16, and ResNet50. Alzheimer disease is a progressive neurodegenerative disorder that affects millions of individuals worldwide. The proposed method aims to leverage the pre-trained weights of these well-established CNN architectures to extract meaningful features from MRI images and improve classification accuracy. Additionally, a novel CNN architecture is introduced, specifically designed for Alzheimer's classification, which incorporates domain-specific knowledge and architectural modifications. Experimental results demonstrate that the transfer learning approach, combined with the proposed CNN, achieves superior performance in Alzheimer disease classification, showcasing the potential of transfer learning in improving early diagnosis and treatment of this debilitating condition.