Clustering of Brain Function Network Based on Attribute and Structural Information and Its Application in Brain...

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Psychiatric Disorders

Background:

• Current brain disease diagnosis relies on subjective symptoms and clinical signs, risking psychiatrist bias.
• Objective diagnostic markers are needed to improve the accuracy and reliability of brain disease identification.

Purpose of the Study:

• To develop and validate a brain network clustering model for objective brain disease identification using resting-state functional magnetic resonance imaging (fMRI).
• To assess the model's performance in distinguishing between Alzheimer's disease, Bipolar disorder, and Schizophrenia without relying on clinical information.

Main Methods:

• Utilized resting-state fMRI data to construct brain networks.
• Employed cosine similarity for attribute similarity and sub-network kernels for structure similarity.
• Integrated both similarity measures into a single matrix for spectral clustering analysis.

Main Results:

• The proposed model achieved high clustering consistency for Schizophrenia (100%) and Bipolar disorder (100%).
• Demonstrated significant improvement in clustering performance for Alzheimer's disease with 60.6% consistency.
• Effectively differentiated between the studied brain diseases based on network attributes and structure.

Conclusions:

• The developed brain network clustering model offers a promising, objective approach for diagnosing brain diseases.
• This method has the potential to reduce diagnostic bias and improve accuracy in psychiatric nosology.
• Resting-state fMRI coupled with advanced clustering techniques provides a powerful tool for understanding and classifying neurological and psychiatric conditions.