The Predictive Values of Changes in Local and Remote Brain Functional Connectivity in Primary Angle-Closure Glaucoma Patients According to Support Vector Machine Analysis

Front Hum Neurosci. 2022 May 19:16:910669. doi: 10.3389/fnhum.2022.910669. eCollection 2022.

Abstract

Purpose: The primary angle-closure glaucoma (PACG) is an irreversible blinding eye disease in the world. Previous neuroimaging studies demonstrated that PACG patients were associated with cerebral changes. However, the effect of optic atrophy on local and remote brain functional connectivity in PACG patients remains unknown.

Materials and methods: In total, 23 patients with PACG and 23 well-matched Health Controls (HCs) were enrolled in our study and underwent resting-state functional magnetic resonance imaging (rs-fMRI) scanning. The regional homogeneity (ReHo) method and functional connectivity (FC) method were used to evaluate the local and remote brain functional connectivity. Moreover, support vector machine (SVM) method was applied to constructing PACG classification model.

Results: Compared with the HC, PACG patients showed increased ReHo values in right cerebellum (CER)_8, left CER_4-5, and right CER_8. In contrast, PACG patients showed decreased ReHo values in the bilateral lingual gyrus (LING)/calcarine (CAL)/superior occipital gyrus (SOG) and right postcentral gyrus (PostCG). The ReHo value exhibited an accuracy of 91.30% and area under curve (AUC) of 0.95 for distinguishing the PACG patients from HC.

Conclusion: Our study demonstrated that the PACG patients showed abnormal ReHo value in the cerebellum, visual cortex, and supplementary motor area, which might be reflect the neurological mechanisms underlying vision loss and eye pain in PACG patients. Moreover, the ReHo values can be used as a useful biomarker for distinguishing the PACG patients from HCs.

Keywords: functional connectivity; functional magnetic resonance imaging; primary angle-closure glaucoma; regional homogeneity; support vector machine.