Altered dynamic large-scale brain networks and combined machine learning in primary angle-closure glaucoma.

Area of Science:

• Neuroscience
• Ophthalmology
• Medical Imaging

Background:

• Primary angle-closure glaucoma (PACG) is a severe, irreversible blinding disease.
• Previous research focused on static brain activity, overlooking dynamic network changes in PACG.
• Understanding PACG's impact on brain network dynamics is crucial for developing interventions.

Purpose of the Study:

• To investigate dynamic functional brain network alterations in PACG patients.
• To compare functional connectivity and temporal metrics between PACG patients and healthy controls.
• To identify potential biomarkers for PACG using machine learning.

Main Methods:

• Resting-state functional magnetic resonance imaging (rs-fMRI) and Independent Component Analysis (ICA) were used.
• Dynamic functional network connectivity (dFNC) states were identified using sliding window and k-means clustering.
• A support vector machine (SVM) model was trained to differentiate PACG patients from controls.

Main Results:

• PACG patients exhibited modified functional connectivity within and between large-scale brain networks.
• Abnormalities in dynamic temporal metrics of functional networks were observed in PACG.
• The SVM model successfully differentiated PACG patients from healthy controls based on connectivity data.

Conclusions:

• PACG significantly impacts the dynamic characteristics of functional brain networks.
• Altered brain network dynamics may play a role in PACG pathogenesis and progression.
• These findings highlight potential therapeutic targets for preserving vision and cognitive function in PACG.