Functional changes in fusional vergence‐related brain areas and correlation with clinical features in intermittent exotropia using functional magnetic resonance imaging

Abstract

To explore the functional changes of the frontal eye field (FEF) and relevant brain regions and its role in the pathogenesis of intermittent exotropia (IXT) children via functional magnetic resonance imaging (fMRI). Twenty‐four IXT children (mean age, 11.83 ± 1.93 years) and 28 normal control (NC) subjects (mean age, 11.11 ± 1.50 years) were recruited. During fMRI scans, the IXT children and NCs were provided with static visual stimuli (to evoke sensory fusion) and dynamic visual stimuli (to evoke motor fusion and vergence eye movements) with binocular disparity. Brain activation in the relevant brain regions and clinical characteristics were evaluated. Group differences of brain activation and brain‐behavior correlations were investigated. For dynamic and static visual disparity relative to no visual disparity, reduced brain activation in the right FEF and right inferior occipital gyrus (IOG), and increased brain activation in the left middle temporal gyrus complex (MT+) were found in the IXT children compared with NCs. Significant positive correlations between the fusional vergence amplitude and the brain activation values were found in the right FEF, right IPL, and left cerebellum in the NC group. Positive correlations between brain activation values and Newcastle Control Scores (NCS) were found in the left MT+ in the IXT group. For dynamic visual disparity relative to static visual disparity, reduced brain activation in the right middle occipital gyrus, left cerebellum, and bilateral IPL was found in the IXT children compared with NCs. Significant positive correlations between brain activation values and the fusional vergence amplitude were found in the right FEF and right cerebellum in the NC group. Negative correlations between brain activation values and NCS were found in the right middle occipital gyrus, right cerebellum, left IPL, and right FEF in the IXT group. These results suggest that the reduced brain activation in the right FEF, left IPL, and cerebellum may play an important role in the pathogenesis of IXT by influencing fusional vergence function. While the increased brain activation in the left MT+ may compensate for this dysfunction in IXT children.

Decreased brain activation in the right frontal eye field, left inferior parietal lobule and cerebellum was found in intermittent exotropia (IXT) children, which may play a role in the pathogenesis of IXT by influencing fusional vergence function. Increased activation in MT+ may compensate for this dysfunction in IXT children.

Abbreviations

ANOVA

analysis of variance

BCVA

best‐corrected visual acuity

FEF

frontal eye field

fMRI

functional magnetic resonance imaging

FOV

field of view

GLM

general linear model

IFG

inferior frontal gyrus

IOG

inferior occipital gyrus

IPL

inferior parietal lobule

IXT

intermittent exotropia

MFG

middle frontal gyrus

MNI

Montreal Neurological Institute

MOG

middle occipital gyrus

MT+

middle temporal gyrus complex

MTG

middle temporal gyrus

NCS

Newcastle Control Score

PD

prism diopters

ReHo

regional homogeneity

ROI

region of interest

SPL

superior parietal lobule

TE

echo time

TR

repetition time

1. INTRODUCTION

Strabismus is one of the most common ocular diseases associated with ocular misalignment. Various reports have indicated that the prevalence of strabismus is ~1.76% among children (Griffith et al., 2016) and ranges from 0.06% to 5.65% in Asia (Schaal et al., 2018). Intermittent exotropia (IXT) is the most common type of strabismus (Audren, 2019), and is present in ~48%–92% of exotropia patients (Mohney & Huffaker, 2003). IXT presents as an outward deviation of the eye that is not constant and can be controlled intermittently. The eyes diverge when one is tired, inattentive, daydreaming, or fixed at distance, leading to impaired binocular vision function (Mohney et al., 2019). Both cosmetic effects and loss of stereoscopic function may negatively affect the psychological well‐being of children and their parents (Mao et al., 2021).

However, the underlying pathological mechanisms of IXT remain unclear. The pathogenesis of IXT is related to changes in fusion function (Raab, 2017). including motor fusion (aligns the two eyes' images globally through vergence eye movements) and sensory fusion (matches the two eyes' images and generates stereoscopic depth perception). Many scholars believe that fusion function plays an important role in eye position control through vergence eye movements (Ding & Levi, 2021), and fusion dysfunction leads to deviation of the eyes and further contributes to IXT (Harwerth et al., 1996). Children with IXT had abnormal fusional vergence function, with lower convergence break points and higher divergence break points (Fu et al., 2015). It has also been reported that fusional convergence function gradually decreases with the prolongation of the course of IXT (Hatt et al., 2011). However, the neural mechanism which leads to abnormal fusion function in patients with IXT remains uncertain.

Functional magnetic resonance imaging (fMRI) is a powerful, noninvasive technique for mapping brain function. Several areas of the human brain have been found to participate in binocular fusion. In the early years, Hasebe et al. (1999) proved the participation of the right fusiform gyrus and left inferior parietal lobule (IPL) in binocular fusion, and Rokers et al. (2009) reported that the middle temporal lobe was activated with the fusion stimulus. In recent years, studies on the localization of fusional vergence centers in humans have gradually deepened, including the frontal eye field (FEF), middle occipital gyrus (MOG), middle temporal gyrus (MTG), and superior parietal lobule (SPL; Li et al., 2016; Peng et al., 2021) found increased activation intensity of bilateral SPL and IPL in IXT patients compared with normal subjects using fMRI, but dynamic visual stimuli which could evoke motor fusion was not applied in this study. Cerebellar lesions were also found to induce vergence deficits (Sander et al., 2009) Clinical manifestations of strabismus may occur when the fusional vergence center appears to change pathologically (Wojtczak‐Kwasniewska et al., 2018). However, functional change of cerebral region involved in fusional function in patients with IXT, and how this leads to the onset of this disease remain unclear.

In this study, we determined whether the FEF and relevant brain regions are related to fusion function, especially to fusional vergence via fMRI. Further, we explored the functional change of these regions in the etiology and pathogenesis of IXT.

2. MATERIALS AND METHODS

2.1. Subjects

The study protocol was approved by the Medical Research Ethics Committee of Beijing Tongren Hospital, Capital Medical University, and written informed consent was obtained from all participants. The methodology adheres to the tenets of the Declaration of Helsinki for research involving human subjects. Twenty‐four children with IXT (14 males, 10 females; mean age, 11.83 ± 1.93 years; age range, 9–15 years) and 28 normal control (NC) subjects (12 males, 16 females; mean age, 11.11 ± 1.50 years; age range, 9–14 years) were recruited at Beijing Tongren Hospital from January 2021 to March 2022. All participants were right‐handed.

The inclusion criteria for children with IXT were as follows: (1) diagnosis of the basic type of IXT (distance deviation was within 10 prism diopters [PD] of the near deviation) based on medical history and clinical examination; (2) age between 9 and 15 years (because children younger than 9 years were too young to cooperate with MRI examination); (3) exotropia deviation between 15 and 80 PD; (4) best corrected visual acuity (BCVA) in each eye was 20/20; (4) no history of ocular or systemic diseases; (5) ability to understand and cooperate with all examinations; and (6) the guardian voluntarily provided written informed consent.

Participants were excluded if they had any of the following: (1) ocular disease (amblyopia, cataracts, glaucoma, optic neuritis, or macular degeneration); (2) combined oblique muscle dysfunction, A sign and V sign, or nystagmus; (3) medium and high hyperopia (spherical equivalent >3.0 diopters [D]), high myopia (spherical equivalent < −6.0 D), and anisometropia (a difference of >1.5 D); (4) history of eye surgery; (5) systemic dysplasia or brain and nervous system diseases; (6) inability to comply with all required inspections; and (7) contraindication to MRI examination (metal in cardiac pacemaker or prosthesis, or previous head or spinal trauma requiring neurosurgery).

2.2. Ophthalmic examinations

All participants underwent detailed ophthalmological examinations, including measurements of BCVA, fundus examination, synoptophore test, alternate cover test, and Newcastle Control Score (NCS) assessment. Refractive error was measured via streak retinoscopy after cycloplegia using 1% cyclopentolate hydrochloride and 1% tropicamide. The angle of deviation was measured using the prism and alternate cover test with accommodative targets at 33 cm for near and 6 m for distance fixation and indicated as PD with exodeviation. The fusional vergence amplitude was detected by synoptophore. The age at the onset of exodeviation in all patients was determined from the medical history provided by the parents.

2.3. Visual stimuli and tasks

A block design paradigm was used in this study. The visual stimuli were displayed on a liquid‐crystal display (LCD) screen that could be seen by the subject through an angled mirror positioned above the subject's eyes. The distance between the LCD screen and the subject's eyes was ~120 cm and the visual angle was ~11° in the static condition. Both eyes were corrected for BCVA with myopic or astigmatic lenses during the MRI scanning. In addition, red and blue lenses were fixed inside the glasses respectively.

Forty‐eight blocks were uniformly organized into two sessions. The visual stimuli contained two types of videos and two types of static pictures, the red or blue stimuli would be vanished through the same color lens (Figure S1). In one video, the red and blue visual animal stimuli moved from the maximum overlap to complete separation. In the other video, the red and blue visual animal stimuli moved from complete separation to the maximum overlap. One static picture contained the maximum overlapped red and blue visual animal stimuli, with static visual disparity. The red and blue visual animal stimuli did not contain a complete animal alone, but constituted a complete animal together through each eye when they overlapped with a small binocular disparity. Therefore, the binocular sensory fusion function would be evoked in the red and blue animal stimuli condition. Dynamic visual stimuli in the videos were expected to evoke motor fusion via vergence eye movements and subjects were asked to press the button once the binocular fusion was broken or restored. The other static picture contained gray animal stimuli with no visual disparity. Each block contained a video or picture lasting 16 s, and a fixation stimulus lasted for 6 s. At the end of the session, a goodbye picture was presented for 2 s. The sequences of the blocks were pseudorandomized. The participants were trained in practice trials that contained both types of stimuli outside the scanner.

2.4. MRI image acquisition

The participants were scanned using a 3.0 T Siemens Prisma MRI scanner (Siemens Healthineers, Erlangen, Germany) with a standard 64 eight‐channel head coil at the Beijing Tongren Hospital. During scanning, foam pads were placed around the head and earplugs were provided for all subjects to minimize head motion and scanner noise, respectively.

Three‐dimensional T1‐weighted anatomical images were collected using a 3D magnetization‐prepared rapid gradient‐echo imaging sequence with the following scan parameters: repetition time (TR), 2000 ms; echo time (TE), 2.25 ms; flip angle, 8°; field of view (FOV), 256 mm × 256 mm; in‐plane image resolution, 1 mm × 1 mm; slice thickness, 1 mm; and 192 continuous sagittal slices. Functional images were obtained with a simultaneous multi‐slice echo planar imaging sequence, and the scan parameters were as follows: TR, 1000 ms; TE, 33 ms; flip angle, 64°; FOV, 208 mm × 180 mm; in‐planar image resolution, 2 mm × 2 mm; matrix size, 104 × 90; slice thickness, 2 mm; slice gap, 0.4 mm; and 60 slices. A posterior‐to‐anterior phase‐encoding direction was used and the scanning axes were parallel to the anterior commissure (AC)–posterior commissure (PC) line.

2.5. Data analysis

For the demographic comparisons, statistical analyses were performed using SPSS Statistics version 22.0 (SPSS Inc., Chicago, Illinois). Age was compared using a two‐sample two‐tailed t‐test and sex was compared using Fisher's exact probability test for gender. Significant difference was set at a level of p < .05.

MRI data were preprocessed and analyzed using statistical parametric mapping software (SPM12; Wellcome Trust Centre for Neuroimaging, London, UK) in MATLAB 2018 (MathWorks, Natick, Massachusetts). Slice timing and head motion correction were then performed for the fMRI images. Subjects' data were excluded from the following analysis if either the translation or rotation of their head motion exceeded 4 mm or 4° on any axis. Functional images were spatially normalized to the Chinese pediatric template in the Montreal Neurological Institute (MNI) standard space. After normalization, functional images were spatially smoothed with a 4 mm full‐width at half‐maximum isotropic Gaussian kernel and then high‐pass filtered with a cutoff frequency of 0.008 Hz to eliminate low‐frequency signal drifts. A general linear model (GLM) analysis was used to determine the cortical regions involved in the processing of different conditions. The onset and duration of the experimental stimuli were convolved with the canonical hemodynamic response function. Head motion parameters were included in the GLM as regressors of movement‐related variance. Then, the parameter estimates for each condition were obtained. Two contrasts were generated: visual disparity contrast defined as red and blue visual stimuli condition (dynamic and static visual disparity condition) > gray visual stimuli condition, and dynamic visual disparity contrast defined as dynamic > static visual stimuli condition.

Analyses of variance (ANOVAs) were used to explore main effect of condition and group on brain activation, as well as their interaction. First, 2 × 2 ANOVAs were performed on cortical activation across the whole brain with group (IXT vs. NC) as the between‐subject factor and condition (dynamic and static visual disparity vs. no visual disparity; dynamic visual disparity vs. static visual disparity) as the within‐subject factor. Individual contrast images were entered into a second‐level one‐sample t‐test analysis to determine the group‐level activation in each group. Significant results were defined as a threshold with a voxel‐wise p < .001 and cluster‐level family‐wise error (FWE) correction at p < .05.

Brain areas related to oculomotor control have been well‐studied in both humans and mammals. Therefore, we selected regions of interest (ROIs) based on activated brain region in both previous studies (Peng et al., 2021) and our study. Next, ROI analyses were adopted to investigate brain activation in each group and the corresponding group differences. Contrast values were extracted from spherical ROIs with a 4 mm radius in contrast images.

Finally, correlation analyses between contrast images and ophthalmologic examinations were performed in both groups to determine the underlying pathophysiological mechanisms related to functional brain changes in children with IXT.

3. RESULTS

3.1. Demographic and clinical features

Six children with IXT and eight NCs were excluded from the analysis because of excessive head movement. The remaining 20 NC subjects and 18 children with IXT were matched in terms of sex, age, and education (IXT: 10 males, 8 females; mean age, 12.06 ± 1.86 years; age range, 9–15 years; NC: 8 males, 12 females; mean age, 11.25 ± 1.41 years; age range, 9–14 years). The clinical characteristics of the enrolled children with IXT and NCs are listed in Table 1.

TABLE 1.

Clinical characteristics of enrolled 18 IXT children and 20 NCs.

	IXT (N = 18)	NC (N = 20)	t‐Value	p‐Value
Sex (male/female)	10/8	8/12	0.920	.338
Age (years)	12.056 ± 1.859	11.250 ± 1.410	1.513	.139
Handedness	18R	20R	N/A	>.999
Education (years)	6.056 ± 1.862	2.500 ± 1.395	1.047	.302
Duration (years)	3.500 ± 2.722	N/A	N/A	N/A
Newcastle control score	6.611 ± 1.037	N/A	N/A	N/A
BCVA (R)	1.061 ± 1.335	1.040 ± 0.821	0.594	.556
BCVA (L)	1.022 ± 0.943	1.040 ± 0.821	−0.621	.538
Spherical equivalent refraction (R)	−2.042 ± 1.366	−1.981 ± 1.355	−0.130	.897
Spherical equivalent refraction (L)	−1.508 ± 1.545	−1.731 ± 1.406	0.445	.659
Near deviation (PD)	−54.444 ± 12.472	N/A	N/A	N/A
Distance deviation (PD)	−49.722 ± 13.114	N/A	N/A	N/A
Fusional vergence amplitude	16.500 ± 4.506	23.400 ± 7.207	−2.822	.014

Note: Data are presented as mean ± SD.

Abbreviations: BCVA, best‐corrected visual acuity; IXT, intermittent exotropia; L, left; N/A, not applicable; NC, normal control; PD, Prism diopters; R, right.

3.2. Brain activation of visual disparity contrast

2 × 2 ANOVAs with group (IXT vs. NC) as the between‐subject factor and condition (dynamic and static visual disparity vs. no visual disparity) as the within subject factor across the whole brain revealed a main effect of condition (p‐voxel <.001, p‐cluster <.05 FWE‐corrected). No significant main effect of group and interaction was found at this threshold.

Relative to the no visual disparity condition (gray stimuli), the dynamic and static disparity conditions (red and blue stimuli) elicited increased activation in the bilateral MFG including FEF, SPL, IPL, IOG, MOG, MTG complex (MT+) and medial frontal gyrus in both groups (p‐voxel <.001, p‐cluster <.05 FWE‐corrected; Figure 1a,b and Table 2). Significant brain activation was found in the bilateral cerebellum in both groups at a lower threshold (p‐voxel <.005, p‐cluster <.05 FWE‐corrected, Figure 1a,b).

FIGURE 1.

Brain activation for visual disparity contrast in intermittent exotropia (IXT) children and normal control (NC). (a) Whole brain activation in the IXT group; (b) Whole brain activation in the NC group; (c) Regions of interest (ROI) analyses of brain activation. The asterisks represent significant interactions between group and condition on brain activation in the right frontal eye field (FEF), right inferior occipital gyrus (IOG) and left middle temporal gyrus complex (MT+; FDR corrected p < .05). IPL, inferior parietal lobule; MOG, middle occipital gyrus.

TABLE 2.

Regions revealing significant brain activation for visual disparity contrast in IXT children and NCs.

Regions	BA	MNI coordinates			Voxels	Z‐value
		x	y	z
IXT
L inferior temporal gyrus	37	−52	−74	−2	541	5.73
L superior parietal lobule		−28	−60	52	505	4.78
L precentral gyrus	6	−48	−4	40	130	4.86
L inferior frontal gyrus		−56	14	18	97	3.99
L insula		−30	24	6	181	4.38
L superior frontal gyrus		−4	12	54	89	3.99
L middle frontal gyrus		−24	−8	54	42	3.75
R inferior parietal lobule	40	36	−56	48	222	5.39
R middle occipital gyrus		44	−72	−8	381	4.89
R superior/middle frontal gyrus		38	38	22	204	3.91
R precentral gyrus		50	8	14	347	4.34
R inferior frontal gyrus		42	4	28	157	4.24
NC
L inferior parietal lobule		−42	−56	54	1037	5.52
L inferior frontal gyrus	9	−46	2	36	822	5.15
L middle frontal gyrus	6	−54	2	42	822	4.84
L sub gyral		−32	26	0	105	5.04
L insula	13	−32	22	10	105	4.57
R superior parietal lobule	7	34	−62	54	971	5.81
R fusiform gyrus	19	38	−68	−18	452	5.74
R inferior parietal lobule	40	36	−56	48	325	5.73
R inferior frontal gyrus	9	52	10	36	632	6.44

Note: p‐voxel <.001, p‐cluster <.05 family‐wise error‐corrected.

Abbreviations: BA, Brodmann area; IXT, intermittent exotropia; L, left; MNI, Montreal Neurological Institute; NC, normal control; R, right.

Based on brain activation, bilateral FEF, IPL, IOG, MT+, and cerebellum were selected as key regions for further ROI analyses, and spherical ROIs with a 4‐mm radius were extracted. 2 × 2 ANOVAs on contrast values of individual ROIs revealed significant main effects of group in the right FEF (p = .002), right IOG (p = .002) and left MT+ (p = .01) and significant interactions in the right FEF (p < .001), right IOG (p = .002), and left MT+ (p = .004; Figure 1c). The IXT group exhibited decreased brain activation in the right FEF and right IOG, and increased brain activation in the left MT+ than NC group, especially for visual disparity condition.

3.3. Brain activation of dynamic visual disparity contrast

2 × 2 ANOVAs with group (IXT vs. NC) as the between‐subject factor and condition (dynamic visual disparity vs. static visual disparity) as the within subject factor across the whole brain revealed main effects of condition. Significant interactions between condition and group were found in the bilateral MOG (p‐voxel <.001, p‐cluster <.05 FWE‐corrected) and bilateral cerebellum (p‐voxel = .005, p‐cluster =.05 FWE‐corrected).

Regarding dynamic disparity conditions relative to the static disparity condition, significant activation was identified in the right FEF, bilateral SPL, IPL, IOG, MOG, and cerebellum in both groups (p‐voxel <.001, p‐cluster <.05 FWE‐corrected, Figure 2a,b and Table 3).

FIGURE 2.

Brain activation for dynamic visual disparity contrast in intermittent exotropia (IXT) children and normal control (NC). (a) Whole brain activation in the IXT group; (b) Whole brain activation in the NC group; (c) Brain activation significantly reduced in the left cerebellum and right middle occipital gyrus (MOG) in the IXT group compared with the NC group (p‐voxel <.001, p‐cluster <.05 family‐wise error‐corrected); (d) Regions of interest analyses of brain activation. The asterisks represent significant interactions between group and condition on brain activation in the bilateral inferior parietal lobule (IPL) and right middle temporal gyrus complex (MT+; FDR corrected p < .05).

TABLE 3.

Regions revealing significant brain activation for dynamic visual disparity contrast in IXT children and NCs.

Regions	BA	MNI coordinates			Voxels	Z‐value
		x	y	z
IXT
L inferior parietal lobule		−38	−40	44	114	3.91
L lingual gyrus		−18	−90	−14	283	5.14
L inferior/middle occipital gyrus	19	−30	−94	4	777	5.64
R middle occipital gyrus		28	−90	2	315	5.55
R inferior parietal lobule		46	−44	46	273	4.75
R middle frontal gyrus		30	2	62	125	3.87
R inferior frontal gyrus		50	12	26	184	3.80
R cerebellum		6	−84	−22	75	3.03
NC
L middle occipital gyrus		−36	−74	−14	441	5.53
L inferior parietal lobule		−42	−56	54	1037	5.52
L inferior frontal gyrus	9	−46	2	36	822	5.15
L middle frontal gyrus	6	−54	2	42	822	4.84
L sub gyral		−32	26	0	105	5.04
L insula	13	−32	22	10	105	4.57
R inferior frontal gyrus	9	52	10	36	632	6.44
R superior parietal lobule	7	34	−62	54	971	5.81
R fusiform gyrus	19	38	−68	−18	452	5.74
R inferior parietal lobule	40	36	−56	48	325	5.73
R cerebellum		10	−84	−24	431	3.35

Note: p‐voxel < .001, p‐cluster < .05 family‐wise error‐corrected.

Abbreviations: BA, Brodmann area; IXT, intermittent exotropia; L, left; MNI, Montreal Neurological Institute; NC, normal control; R, right.

Whole brain comparisons revealed decreased activation in the right MOG and left cerebellum of the IXT group when compared with that of the NC group (p‐voxel <.001, p‐cluster <.05 FWE‐corrected), for dynamic visual disparity contrast (Figure 2c and Table 4).

TABLE 4.

Regions revealing significant brain activation differences between IXT children and NCs for dynamic visual disparity contrast.

Regions	BA	MNI coordinates			Voxels	Z‐value
		x	y	z
IXT < NC
L cerebellum		−16	−72	−18	86	4.07
R cuneus		12	−94	8	77	4.26
R lingual gyrus	18	16	−70	−8	41	3.75
R fusiform gyrus		28	−66	−8	33	3.70

Note: p‐voxel <.001, p‐cluster <.05 family‐wise error‐corrected.

Abbreviations: BA, Brodmann area; IXT, intermittent exotropia; L, left; MNI, Montreal Neurological Institute; NC, normal control; R, right.

Brain activation for this contrast in the ROIs adopted in the previous section was extracted. 2 × 2 ANOVAs on contrast values of individual ROIs revealed significant main effects of group in the right MT+ (p = .002) and significant interactions in the left IPL (p = .027), right IPL (p = .02), and right MT+ (p < .001). The IXT group exhibited reduced brain activation than NC in the right MT+, especially for dynamic visual disparity condition. Compared with the NC group, the IXT group exhibited reduced brain activation for dynamic disparity conditions relative to static disparity condition in the bilateral IPL (Figure 2d).

3.4. Brain‐behavior analyses of visual disparity contrast

Brain–behavior analyses revealed significant positive correlations between the fusional vergence amplitude and brain activation values for the visual disparity contrast in the right FEF (r = .53, p = .017), right IPL (r = .54, p = .014), and left cerebellum (r = .56, p = .0096) in the NC group (Figure 3a).

FIGURE 3.

Correlations between the behavioral measurements and brain activation for visual disparity contrast. (a) Brain activation in the right frontal eye field (FEF), right inferior parietal lobule (IPL), and left cerebellum was positively correlated with fusional vergence amplitude in the normal control (NC); (b) Brain activation in the left middle temporal gyrus complex (MT+) was positively correlated with the NCS scores in intermittent exotropia children.

Contrast values for the visual disparity contrast in the left MT+ were positively correlated with NCS (Spearman rank correlation: r = .64, p = .004) in the IXT group (Figure 3b).

3.5. Brain‐behavior analyses of dynamic visual disparity contrast

Significant positive correlations between brain activation values for dynamic disparity conditions relative to static disparity condition and the fusional vergence amplitude were found in the right FEF (r = .5, p = .026) and right cerebellum (r = .7, p < .001) in the NC group (Figure 4a).

FIGURE 4.

Correlations between the behavioral measurements and brain activation for dynamic disparity contrast. (a) Brain activation in the right frontal eye field (FEF) and right cerebellum was positively correlated with fusional vergence amplitude in the normal control (NC); (b) Brain activation in the middle occipital gyrus (MOG), right cerebellum, left inferior parietal lobule (IPL), and right FEF was negatively correlated with the Newcastle Control Scores (NCS) scores in intermittent exotropia (IXT) children.

Negative correlations between NCS and contrast values for dynamic visual disparity contrast were found in the right MOG (Spearman rank correlation: r = −.56, p = .016), right cerebellum (r = −.61, p = .0068), left IPL (r = −.62, p = .006) and right FEF (r = −.54, p = .02) in the IXT group (Figure 4b).

4. DISCUSSION

In this study, activation changes in the FEF and related brain regions under binocular visual stimuli and its relationship with clinical features in children with IXT were investigated using fMRI. Our findings help to explore the role of fusional function, especially fusional vergence in the pathogenesis of IXT.

4.1. Brain activation in IXT children and NCs

Several areas were found to participate in binocular fusion. Li et al. (2016) proved that the programming of binocular fusion involved a brain network covering several regions including the MOG, SPL, and IPL, which was also found in our study. Under dynamic and static visual disparity condition, more extensive brain regions, including bilateral FEF, IOG, MT+, medial frontal gyrus, and cerebellum showed increased activation for binocular fusion in both group in our study.

Few studies have designed dynamic disparity stimulation tasks to investigate fusional vergence function. We devised dynamic stimulation which evoke motor fusion to investigate the change on brain activation only for fusional vergence, which plays a more important role in the pathogenesis of IXT. Under dynamic disparity condition versus static disparity condition, significant activation was found in the right FEF, bilateral SPL, IPL, IOG, MOG, and cerebellum in both group, suggesting their roles especially in fusional vergence.

4.2. Differences in the activation values of key regions between IXT and NC groups

In the visual disparity contrast, the ROI analyses showed reduced activation in the right FEF, right IOG and increased activation in the left MT+ in children with IXT compared with NCs. The FEF has been found to play a role in vergence eye movement in the primate frontal cortex in animal experiments (Kuang et al., 2016). In the study conducted by Morales et al. (2020) functional activation within the FEF was found to be increased during vergence eye movement compared with that during rest. We speculate that the decreased activation in the FEF may reflect dysfunctional binocular fusion in children with IXT. As a part of ventral visual stream, IOG is correlated with spatial and face‐part selectivity (de Haas et al., 2021). The right IOG made plausible causal contributions to visuospatial attention (Toba et al., 2017). The decreased activation in the right IOG in our study may reflect potential impairment of spatial attention and selectivity in children with IXT. The MT+ is known to play a role in high‐cognitive functions, including vision, eye fixation, and pursuit movement (Kwan et al., 2021). In children with IXT, more effort may be required for fixation and pursuit to maintain a single image with two eyes. We speculated that the increased activation in the left MT+ might reflect compensation for the damage of fusional vergence function in children with IXT.

In the dynamic visual disparity contrast, whole brain comparisons revealed decreased activation in the left cerebellum and right MOG of the IXT group compared with NC group. Clinical observations indicate that cerebellar lesions may lead to vergence deficits (Sander et al., 2009). The ocular motor vermis, caudal fastigial nuclei, and posterior interposed nuclei in cerebellum are involved in vergence eye movements (Beh et al., 2017). Purkinje cells in the cerebellar dorsal vermis in primate were confirmed to be involved in vergence pursuit (Nitta et al., 2008). The decreased activation in the left cerebellum of the IXT group may contribute to the deviation of eye position by affecting vergence eye movements. Brain activation related to visual perception was mainly concentrated in the occipital cortex, while the early processing of visual awareness activation was located in the MOG (Tu et al., 2013). Activation of the MOG is considered as a part of the visual dorsal stream in normal subjects (Wandell et al., 2007). Mon‐Williams et al. (2001) found that the dorsal stream has access to binocular information (horizontal image disparities and vergence). The decreased activation in the right MOG of the IXT group may indicate the abnormality of visual dorsal stream and binocular fusion in IXT children.

In the dynamic visual disparity contrast, the ROI analyses revealed reduced activation in bilateral IPL in the IXT group compared with the NC group. Previous study also found reduced functional connectivity values in the IPL in adults with strabismus (Liu et al., 2022). The main visual pathways include the ventral and dorsal streams (Lee et al., 2000). Visual information processed through the dorsal visual pathway reaches the IPL (Kobayashi, 2016), and this pathway primarily participates in spatial position information processing and eye movement (Barany et al., 2020). Hasebe et al. (1999) confirmed that the left IPL participates in binocular fusion. The reduced activation of bilateral IPL may be associated with the abnormal fusion function in the IXT group.

4.3. Correlation between activation in related brain areas and behavioral characteristic in the NC and IXT groups

In the NC group, brain activation for visual disparity contrast in the right FEF, right IPL, and left cerebellum, and brain activation for dynamic disparity contrast in the right FEF and right cerebellum was found to be positively correlated with fusional vergence amplitude, which reflects fusional vergence function. The eye position is controlled by the fusion mechanism through vergence eye movement, and the pathogenesis of IXT is associated with abnormal fusion function (Raab, 2017). Our previous study revealed a smaller fusional vergence amplitude of children with IXT than NC. Children with IXT have an abnormal fusional function, with a decreased range of convergence fusion and an increased range of divergence fusion (Fu et al., 2015). This result implies that brain activation in the right FEF, right IPL, and cerebellum may be associated with fusional vergence function, therefore abnormal activation in these regions may participates in the pathogenesis of IXT.

Furthermore, we investigated the correlations between activation in related brain areas and NCS in the IXT group. The NCS is commonly used to evaluate the eye position control ability in IXT patients, which is a significant way to evaluate the severity of IXT. Higher NCS represents a weaker control of eye position and more severe IXT. Brain activation for dynamic disparity contrast in the right MOG, right cerebellum, left IPL, and right FEF was found to be negatively correlated with NCS, suggesting their roles in the pathogenesis of IXT by impaired eye position control. In addition, brain activation for visual disparity contrast in the left MT+ was found to be positively correlated with NCS in the IXT group. This result indicates a link between more severe IXT and greater compensatory activation in the left MT+.

4.4. Study limitations

This study had some inevitable limitations. First, the sample size was relative small because the subjects are limited to children due to age of onset of IXT, especially for us to recruit NC children. Second, the cooperation efficiency is low during testing. Because of the subject's head movement, some data could not be used, which is another reason for the limited sample size. We will appropriately increase sample size in future studies. Third, only the brain activation of IXT was investigated. It remains uncertain whether the results can be applied to other forms of concomitant exotropia.

5. CONCLUSIONS

In this study, we investigated brain activation of IXT children and NCs for dynamic and static visual disparity and identified relative brain regions associated with fusion function and vergence eye movement. Reduced brain activation in the right FEF, left IPL, and cerebellum was found in IXT children, which might play an important role in the pathogenesis of IXT by influencing fusional vergence function, while increased brain activation in the left MT+ might compensate for this dysfunction in IXT children.

CONFLICT OF INTEREST STATEMENT

None of the authors has any conflicts of interest to disclose.

Supporting information

FIGURE S1. (A) Condition 1: Dynamic visual animal stimuli 1 in which the red and blue visual animal stimuli moved from the maximum overlap to complete separation; (B) Condition 2: Dynamic visual animal stimuli 2 in which the red and blue visual animal stimuli moved from complete separation to the maximum overlap; (C) Condition 3: Red and blue visual animal stimuli with static visual disparity; (D) Condition 4: Gray visual animal stimuli with no visual disparity.

Zhang, W. , Fei, N. , Wang, Y. , Yang, B. , Liu, Z. , Cheng, L. , Li, J. , Xian, J. , & Fu, T. (2023). Functional changes in fusional vergence‐related brain areas and correlation with clinical features in intermittent exotropia using functional magnetic resonance imaging. Human Brain Mapping, 44(15), 5002–5012. 10.1002/hbm.26427

DATA AVAILABILITY STATEMENT

The data that have been used are confidential.