Hybrid-RViT: Hybridizing ResNet-50 and Vision Transformer for Enhanced Alzheimer's disease detection

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Neurology

Background:

• Alzheimer's disease (AD) is a major global health concern, necessitating early detection for effective management.
• Current diagnostic methods for AD can be improved with advanced analytical tools.

Purpose of the Study:

• To develop and evaluate a novel deep learning (DL) model, Hybrid-RViT, for enhanced detection and classification of Alzheimer's disease stages.
• To leverage a hybrid approach combining Convolutional Neural Networks and Vision Transformers for improved feature extraction from brain MRI data.

Main Methods:

• The Hybrid-RViT model was developed by integrating ResNet-50 for feature extraction and Vision Transformer (ViT) for capturing long-range dependencies in brain MRI images.
• Transfer learning using ResNet-50 was employed to enhance inductive bias and feature representation.
• The ViT component processed image patches using self-attention mechanisms for joint local-global feature extraction.

Main Results:

• The Hybrid-RViT model achieved a high training accuracy of 97% and a testing accuracy of 95%.
• The model demonstrated superior performance compared to existing methods in classifying Alzheimer's disease stages from MRI data.
• The Hybrid-RViT model effectively identified and classified different stages of AD.

Conclusions:

• The Hybrid-RViT model shows significant potential as a valuable tool for medical professionals in analyzing brain MRI images for Alzheimer's disease detection.
• This deep learning approach can substantially improve the accuracy and efficiency of early AD diagnosis and intervention.
• The hybrid architecture offers a promising direction for advancing AI-driven diagnostic tools in neurodegenerative disease research.