Machine learning analysis reveals abnormal functional network hubs in the primary angle-closure glaucoma patients

Area of Science:

• Neuroscience
• Ophthalmology
• Medical Imaging

Background:

• Primary angle-closure glaucoma (PACG) is a severe, irreversible blinding eye condition.
• PACG is associated with vision and vision-related brain region alterations.
• Whole-brain functional network hub changes in PACG remain largely unexplored.

Purpose of the Study:

• To investigate brain functional network hub alterations in PACG patients.
• Utilized voxel-wise degree centrality (DC) to analyze brain network hubs.

Main Methods:

• 31 PACG patients and 31 healthy controls (HCs) were enrolled.
• Voxel-wise degree centrality (DC) method applied to assess brain function.
• Support vector machine (SVM) used for PACG patient classification.

Main Results:

• PACG patients showed higher DC in the right fusiform, left middle temporal gyrus, and left cerebellum.
• Lower DC values observed in PACG patients in the right calcarine, right postcentral gyrus, left precuneus, and left postcentral gyrus.
• SVM achieved 72.58% accuracy with an AUC of 0.85 for classifying PACG patients.

Conclusions:

• PACG patients exhibit widespread brain functional network hub dysfunction.
• Alterations span visual, auditory, default mode, and cerebellum networks.
• Findings offer insights into the neural mechanisms of optic atrophy in PACG.