A Novel Graph Neural Network Framework for Brain Age Prediction.

Area of Science:

• Neuroscience
• Artificial Intelligence
• Medical Imaging

Background:

• Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by cognitive decline.
• Early detection of AD is crucial for timely intervention but remains a significant challenge.
• Resting-state functional MRI (rs-fMRI) offers insights into brain connectivity patterns relevant to neurodegeneration.

Purpose of the Study:

• To develop and evaluate a novel deep learning model for brain age prediction using rs-fMRI data.
• To assess the model's efficacy in identifying early signs of Alzheimer's disease.
• To improve the accuracy of brain age estimation for neurodegenerative disease research.

Main Methods:

• Proposed a Hierarchical GCN-Transformer Network (HGTNet) integrating Graph Convolutional Networks (GCN) and Transformer architectures.
• Utilized rs-fMRI data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset for model training and validation.
• Compared HGTNet performance against traditional machine learning and other deep learning approaches.

Main Results:

• HGTNet demonstrated superior performance in brain age prediction compared to existing methods.
• The integrated GCN and Transformer architecture effectively captured complex brain interactions from rs-fMRI data.
• The model achieved more accurate brain age predictions, indicating its potential for early AD detection.

Conclusions:

• The developed HGTNet model shows significant promise for early detection of neurodegenerative changes associated with Alzheimer's disease.
• Accurate brain age prediction using rs-fMRI can serve as a valuable biomarker for AD.
• This approach facilitates better intervention and management strategies for Alzheimer's disease.