Abnormal Low-Frequency Oscillations Reflect Abnormal Eye Movement and Stereovision in Patients With Comitant Exotropia

Abstract

Background: Patients with comitant exotropia (CE) are accompanied by abnormal eye movements and stereovision. However, the neurophysiological mechanism of impaired eye movements and stereovision in patient with CE is still unclear. Purpose: The purpose of this study is to investigate spontaneous neural activity changes in patients with CE using the amplitude of low-frequency fluctuation (ALFF) method and the machine learning method. Materials and Methods: A total of 21 patients with CE and 21 healthy controls (HCs) underwent resting-state magnetic resonance imaging scans. The ALFF and fractional amplitude of low-frequency fluctuation (fALFF) values were chosen as classification features using a machine learning method. Results: Compared with the HC group, patients with CE had significantly decreased ALFF values in the right angular (ANG)/middle occipital gyrus (MOG)/middle temporal gyrus (MTG) and bilateral supplementary motor area (SMA)/precentral gyrus (PreCG). Meanwhile, patients with CE showed significantly increased fALFF values in the left putamen (PUT) and decreased fALFF values in the right ANG/MOG. Moreover, patients with CE showed a decreased functional connectivity (FC) between the right ANG/MOG/MTG and the bilateral calcarine (CAL)/lingual (LING) and increased FC between the left PUT and the bilateral cerebellum 8/9 (CER 8/9). The support vector machine (SVM) classification reaches a total accuracy of 93 and 90% and the area under the curve (AUC) of 0.93 and 0.90 based on ALFF and fALFF values, respectively. Conclusion: Our result highlights that patients with CE had abnormal brain neural activities including MOG and supplementary motor area/PreCG, which might reflect the neural mechanism of eye movements and stereovision dysfunction in patients with CE. Moreover, ALFF and fALFF could be sensitive biomarkers for distinguishing patients with CE from HCs.

Keywords: ALFF; concomitant exotropia; fALFF; functional magnetic resonance imaging; machine learning; support vector machine.