Sage Journals: Discover world-class research

Abstract

Background

The quantitative assessment of the oculomotor system has emerged as a promising biomarker for neurodegenerative disorders. Although oculomotor impairments are commonly observed in multiple system atrophy (MSA) patients, the specific abnormalities and underlying neural structural changes remain poorly understood.

Objectives

To explore oculomotor abnormalities and associated brain changes in MSA, evaluating their potential as biomarkers for diagnosis and disease monitoring.

Methods

A total of 100 MSA patients and 50 healthy controls (HCs) were included in this study. All subjects underwent comprehensive evaluations, including clinical assessments, virtual reality (VR)-based ocular-tracking tasks and structural magnetic resonance imaging (MRI).

Results

Compared with HCs, MSA patients showed significantly impaired smooth pursuit (SP) with increased number of deviations (13.33 [29.33] vs. 5.67 [7.67], p < 0.001); reduced prosaccadic (PS) average velocity (194.80 ± 82.45 °/s vs. 263.07 ± 68.17 °/s, p < 0.001); and reduced antisaccade (AS) average velocity (165.82 ± 85.75 °/s vs. 257.05 ± 74.39 °/s, p < 0.001). A combination of PS and AS average velocities with SP number of deviations effectively distinguished MSA patients from HCs with an AUC of 0.814. PS average velocity was negatively correlated with UMSARS total scores (r = -0.354, p < 0.001), whereas AS accuracy was positively correlated with MoCA scores (r = 0.375, p = 0.001). Voxel-based morphometry revealed significant associations between these oculomotor parameters and atrophy in the cerebellum and frontal gyrus (p < 0.05, family-wise error correction).

Conclusions

Our study provides comprehensive insights into the VR-based quantitative oculomotor analysis and its association with regional brain atrophy in MSA, contributing to novel biomarkers identification and therapeutic targets exploration.

Plain Language Summary Title

Using Virtual Reality to Measure Eye Movement Problems and Brain Shrinkage in Multiple System Atrophy

Plain Language Summary

What is the research about? Multiple system atrophy (MSA) is a rare, progressive brain disease that causes problems with movement, balance, and other body functions. People with MSA often have trouble with eye movements, but we don't fully understand these issues or how they relate to brain changes. This study aimed to use a new virtual reality (VR) tool to measure eye movements and link them to brain shrinkage, to find better ways to diagnose and track the disease. Why is this important? MSA is hard to diagnose early because symptoms overlap with other conditions like Parkinson's disease. Current tests rely on doctors' observations, which aren't always precise. Eye movements involve many brain areas affected by MSA, so measuring them accurately could provide objective “biomarkers” for earlier detection and monitoring disease progression. How was the study done? We studied 100 people with MSA and 50 healthy volunteers. Everyone had clinical check-ups, VR-based eye-tracking tests (measuring how eyes fixate, follow moving objects, and make quick jumps), and brain MRI scans to measure brain volume in key areas. What did we find? People with MSA had worse eye movements than healthy volunteers, especially in following objects smoothly and making voluntary or intentional eye jumps. These problems helped distinguish MSA from healthy people. The eye issues were linked to disease severity, and corresponded to shrinkage in brain areas like the cerebellum and frontal regions. What does this mean? VR eye-tracking offers a simple, portable way to spot MSA changes that match real-life behaviors, unlike traditional tests. This could lead to better biomarkers for diagnosing MSA early and testing new treatments, improving care for patients. Future studies should compare MSA to similar diseases and track changes over time.

Keywords

multiple system atrophy oculomotor deficits eye movements structural MRI biomarker

Introduction

Multiple system atrophy (MSA) is a progressive neurodegenerative disorder characterized by widespread neuronal loss and gliosis in the central nervous system, particularly affecting striatonigral, olivopontocerebellar, and central autonomic structures.^1–3 Clinically, MSA presents a diverse spectrum of progressive autonomic, parkinsonian, cerebellar, and pyramidal symptoms.⁴ It is classified into two subtypes based on the predominant motor phenotype: parkinsonian (MSA-P) and cerebellar (MSA-C). However, the overlap and variability of these clinical features pose significant challenges for early diagnosis.⁵ Current diagnostic approaches rely heavily on clinical observations and patient-reported outcomes, which often lack the sensitivity and specificity needed to monitor disease progression or assess therapeutic efficacy in clinical trials.^6–8 Consequently, identifying reliable and quantifiable biomarkers for MSA is critical for advancing future research.⁹

Oculomotor assessment is an essential part of the neurological examination, as it involves neural circuits spanning the cerebrum, brainstem, and cerebellum.¹⁰ In MSA, typical abnormal oculomotor features include gaze-induced nystagmus, mild to moderate hypometria of saccades, impaired smooth pursuit (SP) eye movements with decreased gain, and excessive square-wave jerks.^11,12 Among these, saccadic abnormalities are a highly prevalent impairment in patients with MSA and are considered to hold potential as diagnostic biomarkers. The antisaccade (AS) task, which requires suppression of reflexive responses, largely depends on frontostriatal circuits, and their modulation by the superior colliculus and cerebellum.^13,14 This task has been shown to correlate with disease progression in Parkinson's disease (PD), where nigrostriatal deficits disrupt frontostriatal circuits.^15,16 Given MSA involves more extensive neurodegeneration in these pathways compared to PD, similar AS impairments may reflect MSA progression. However, these oculomotor impairments have been scarcely studied in MSA.

Recent virtual reality (VR)- based quantitative eye movement assessments present several advantages over traditional approaches.¹⁷ First, this approach enables sensitive detection of clinically significant oculomotor signatures with high objectivity and reproducibility, independent of clinician experience required for bedside assessment.¹⁸ Moreover, VR delivers stimuli in an immersive three-dimensional environment that minimizes external interference and elicits natural gaze behaviors, providing a closer approximation to real-world dynamics than the static screens used in traditional quantitative assessments. Additionally, the intuitive interface and portability of VR systems facilitate efficient and accessible operation in diverse settings.¹⁹

Despite these advantages, applications to MSA remain unexplored. Moreover, growing evidence suggests that oculomotor parameters reflect specific disruptions within affected cerebral networks.^20,21 Combining oculomotor analysis with neuroimaging could help clarify the anatomical substrates of these abnormalities and provide deeper insights into their neural mechanisms. However, current evidence remains limited by small cohort sizes and restricted oculomotor paradigms in prior studies.^21,22

To bridge these gaps, our study comprehensively investigated oculomotor alterations and their neural underpinnings in a large MSA cohort, combining VR-based eye-tracking systems, clinical assessments and structural MRI. We aimed to evaluate their potential as reliable biomarkers.

Methods

Participants study

In this cross-sectional study, we recruited 100 patients with MSA (60 MSA-C and 40 MSA-P) from the Department of Neurology, Xiangya Hospital of Central South University between February 2023 and February 2025. The inclusion criteria required (1) independent availability of complete eye-tracking assessment; (2) a diagnosis of clinically established or clinically probable MSA according to the 2022 Movement Disorder Society Criteria,²³ which include autonomic failure, parkinsonism, or cerebellar ataxia in conjunction with supportive clinical and imaging findings. Diagnoses were established by two experienced neurologists, with an independent blind diagnosis from a third movement disorders expert included in cases of discrepancy. Spinocerebellar ataxia 1, 2, 3, 6, 7, 17 and dentatorubral-pallidoluysian atrophy (DRPLA) were genetically excluded. Additionally, 50 gender- and age-matched healthy controls (HCs) were recruited from patients’ family members and the Physical Examination Center at Xiangya Hospital. All HCs, who had no history of neurological disorders and were able to independently complete the full eye-tracking assessment, underwent standardized screening including review of medical history, basic neurological and imaging examinations.

The exclusion criteria for all participants were as follows: the history of epilepsy, stroke, brain surgery, or psychiatric disease (e.g., schizophrenia, bipolar disorder), unrelated intracranial organic lesions; peripheral vestibulopathy and ophthalmic diseases; other systemic conditions or diseases that could have caused ocular symptoms affecting eye tracking; the history of alcohol or drug abuse; MRI contraindications.

The study received approval from the Ethics Committee of Xiangya Hospital of Central South University, and all participants provided informed consent before their inclusion in the study.

Clinical assessments

All patients underwent clinical assessments by two experienced neurologists. Disease severity was assessed using the Hoehn and Yahr Parkinson's disease staging scheme (H&Y stage) and the Unified Multiple System Atrophy Rating Scale (UMSARS), which includes activities of daily life (UMSARS-I), motor symptoms (UMSARS-II), and chore-based disability (UMSARS-IV).²⁴ Cognitive function was comprehensively evaluated using the Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA) for global cognitive performance,²⁵ while the Frontal Assessment Battery was used for the assessment of frontal executive function (FAB).²⁶

Eye movement recording

Eye movements were recorded binocularly using the National Medical Products Administration-approved eye-tracking assessment system (EyeKnow; Beijing CAS-Ruiyi Information Technology Co., Ltd). The system comprises a tablet computer and a VR headset featuring an eye-tracking camera, with a sampling rate of 120 Hz. The VR headset serves to guide the subject's eye movements and record oculomotor data. Prior to testing, participants completed a nine-point calibration procedure to achieve a maximum calibration error no greater than 2° radius. Participants followed a predesigned protocol to assess central fixation, SP, prosaccade (PS), and AS. Data were processed in real-time using the system's integrated processing module (the detailed procedures are described in the supplementary materials).

The EyeKnow system has been rigorously validated throughout the registration and approval process to ensure their feasibility, reliability, stability, precision, and safety. The utility of these systems has also been confirmed in previous publications.^27,28

Image acquisition and data processing

All participants underwent brain MRI scans using a 3.0 T MRI scanner (Discovery CT750 HD, GE Healthcare). High-resolution T1-weighted images were acquired with the following parameters: number of slices=190; repetition time (TR) = 8.5 ms; echo time (TE) = 3.2 ms; flip angle=12°; field of view=256 × 256 mm; voxel size=1 × 1 × 1 mm; and slice thickness=1.0 mm. Total scan duration was 4 min and 43 s. MRI scans in DICOM format were converted to NIFTI using MRIcro software and processed with the CAT12 toolbox in SPM12 (Wellcome Trust Center for Neuroimaging) via MATLAB R2022a (MathWorks, Natick, MA, USA). Voxel-based morphometry (VBM) analysis was performed, involving image segmentation and spatial normalization using the DARTEL method to generate probability maps for gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF).²⁹ Image quality was assessed through manual visual inspection and SPM's quality metrics. Gray matter volume (GMV) images were smoothed with an 8 mm Gaussian filter. All these tasks were performed in adherence to the protocols outlined in the Cat12 manual.

Statistical analysis

Statistical analyses were performed using IBM SPSS Statistics 31.0 (SPSS Inc.), with a significance threshold of p < 0.05 (two-tailed). The Shapiro-Wilk and Levene tests were applied to confirm the normal distribution of data and homogeneity of variances, respectively. Continuous data were presented as mean ± standard deviation (SD) or median (interquartile range), with parametric data analyzed using t-tests. Categorical data were expressed as counts (%) and assessed via chi-squared (χ2) tests. Non-parametric data analyses were carried out using Mann-Whitney U tests. Oculomotor parameters across groups were compared using t-tests or Mann-Whitney U tests. Diagnostic accuracy was evaluated using receiver operating characteristic (ROC) curves, with optimal thresholds determined by the Youden index and area under the curve (AUC) values compared using the DeLong method. Correlations between oculomotor parameters and clinical variables were examined using Spearman tests. Multiple regression, adjusting for age, gender, and total intracranial volume (TIV), was employed to investigate associations between oculomotor parameters and GMV in brain regions showing significant group differences (p < 0.05, family-wise error correction).

Results

Demographic and clinical characteristics

Demographic and clinical data are presented in Table 1. The study comprised 150 subjects: 100 MSA patients and 50 HCs. No significant differences were observed in age, sex, body mass index (BMI), or years of education between MSA patients and HCs. Among MSA patients, 56 (56%) were male, with a median age of 55.00 (10.0) years. For HCs, 23 (46%) were male, with a median age of 59.00 (10.0) years.

Table 1.

Demographic and clinical characteristics.

	HC (n = 50)	MSA (n = 100)	p
Age (years)*	59.00 (10.0)	55.00 (10.0)	0.297 ^a
Gender (M/F)			0.248 ^b
Male (%)	23 (46%)	56 (56%)
Female (%)	27 (54%)	44 (44%)
BMI^#	23.70 ± 2.93	24.05 ± 3.11	0.517 ^c
Education (years)*	9.00 (6.00)	9.00 (6.00)	0.947 ^a
Age at onset (years)	-	53.13 ± 6.71	-
Disease duration (years)	-	2.25 (2.88)	-
H&Y stage	-	3.00 (0.50)	-
UMSARS total	-	35.00 (23.00)	-
UMSARS-I	-	19.22 ± 7.07	-
UMSARS-II	-	18.00 (11.00)	-
UMSARS-III	-	2.00 (2.00)	-
MMSE (n = 95)	-	27.00 (4.00)	-
MoCA (n = 74)	-	22.59 ± 4.80	-
FAB (n = 88)	-	14.00 (5.00)	-

Note. — ^#: mean ± sd, *: median (interquartile range), BMI: body mass index, H&Y stage: Hoehn and Yahr scale, UMSARS: unified multiple system atrophy rating scale, MMSE: mini-mental state examination, MoCA: Montreal cognitive assessment, FAB: frontal assessment battery. a: Mann-Whitney U test, b: Chi-square test, c: t test.

The group differences of oculomotor parameters

Figure 1 illustrates examples of the four eye movement tasks for a typical MSA and HC. In summary, MSA patients exhibited inferior performance compared to HCs across four aspects of oculomotor function (Table 2). The detailed analysis results of these four aspects are as follows.

Figure 1.

Examples of the four eye movement tasks for a typical MSA and a typical HC. (a) In the fixation task, planar gaze trajectories for the MSA group showed more frequent deviations from the target. (b) The trajectories demonstrate the position changes of the target point (gray line) and participant gaze position (red line). In the smooth pursuit (SP) task, trajectories of the MSA group's paths were disorganized, characterized by frequent catch-up saccades. (c) The horizontal prosaccade (PS) task analysis showed that the MSA group had slower saccades and exhibited saccadic hypometria. (d) In the vertical antisaccade (AS) task, the MSA group demonstrated a higher error rate and extended correction reaction time.

Table 2.

Oculomotor characteristics of HC and MSA.

	MSA (n = 100)	HC (n = 50)	t/Z	p
Fixation
Accuracy (%) *	94.66 (6.81)	95.72 (4.67)	−1.840	0.066 ^a
Number of deviations (>4°) (times) *	2.00 (5.00)	3.00 (4.00)	−0.188	0.851 ^a
Number of deviations (>2°) (times) *	10.50 (17.75)	6.00 (11.00)	−1.714	0.087 ^a
Total deviation (>4°) (°) *	19.10 (29.35)	19.80 (26.26)	−0.304	0.761 ^a
Total deviation duration (ms) *	499.30 (443.37)	388.00 (363.33)	−2.418	0.016 ^a
Smooth Pursuit
Initial time (ms) *	515.79 (279.22)	493.86 (157.78)	−0.975	0.330 ^a
Tracking velocity (°/s) *	21.07 (14.57)	27.41 (20.90)	−2.542	0.011 ^a
Tracking acceleration (°/s²) *	61.26 (85.43)	77.20 (114.96)	−2.304	0.021^a
Number of deviations (times) *	13.33 (29.33)	5.67 (7.67)	−5.353	<0.001 ^a
Deviation (>4°) (°) *	5.02 (0.71)	5.17 (2.73)	−0.584	0.559 ^a
Total deviation (>4°) (°) *	70.50 (153.13)	28.70 (44.92)	−5.182	<0.001 ^a
Prosaccade
Accuracy (%)*	100.00 (0.00)	100.00 (0.00)	−2.242	0.025 ^a
Latency (ms) *	367.78 (107.30)	326.09 (88.22)	−2.803	0.005 ^a
Reaction Time_MIN (ms) *	332.94 (110.91)	266.97 (66.95)	−3.980	<0.001 ^a
Reaction Time_AVG (ms) ^#	482.78 ± 100.36	405.10 ± 74.01	−5.305	<0.001 ^b
Velocity_AVG (°/s) ^#	194.80 ± 82.45	263.07 ± 68.17	5.328	<0.001 ^b
Velocity_MAX (°/s) *	451.38 (262.48)	410.77(232.99)	−0.159	0.873 ^a
Antisaccade
Accuracy (%) *	38.05 (60.63)	52.25 (46.53)	−2.563	0.010 ^a
Latency (ms) ^#	389.32 ± 107.67	361.56 ± 80.90	−1.746	0.083 ^b
Reaction Time_MIN (ms) *	333.70 (99.54)	290.15 (88.73)	−3.337	0.001 ^a
Reaction Time_AVG (ms) *	533.80 (199.14)	453.97 (147.11)	−4.297	<0.001 ^a
Velocity_AVG (°/s) ^#	165.82 ± 85.75	257.05 ± 74.39	6.242	<0.001 ^b
Velocity_MAX (°/s) *	326.29 (258.80)	488.02 (254.72)	−3.805	<0.001 ^a
Error correction rate (%) *	64.59 (69.00)	100.00 (50.00)	−2.334	0.020 ^a
Error correction reaction time (ms) ^#	450.25 ± 158.99	372.09 ± 117.10	−3.017	0.003 ^b

Note. — ^#: mean ± sd, *: median (interquartile range), AVG: average, MIN: minimum, MAX: maximum, a: Mann-Whitney U test, b: t test.

Fixation stability

Patients with MSA exhibited significantly longer total deviation duration than HCs (p = 0.016). No significant differences were observed in accuracy (p = 0.066), number of deviations (>4°) (p = 0.851), number of deviations (>2°) (p = 0.087) or total deviation (>4°) (p = 0.761).

Smooth pursuit

The MSA group showed significantly SP impairment with decreased tracking velocity (p = 0.011) and acceleration (p = 0.021) versus HCs. Both number of deviations (p < 0.001) and total deviation (p < 0.001) were significantly higher in MSA patients. Initial time (p = 0.330) and deviation (>4°) (p = 0.559) did not differ between groups.

Prosaccade

MSA patients exhibited reduced accuracy (p = 0.025) and longer latency (p = 0.005) during target-directed saccades compared to controls. Additionally, the reaction time of PS for the MSA group was significantly slower than HCs, both average (p < 0.001) and minimum (p < 0.001). Average saccadic velocity was significantly reduced in MSA patients (p < 0.001), but maximum saccadic velocity showed no significant difference (p = 0.873).

Antisaccade

The MSA group had significantly lower AS accuracy (p = 0.010) and slower reaction times, both average (p < 0.001) and minimum (p = 0.001) compared to HCs. Both average (p < 0.001) and maximum (p < 0.001) AS velocities were significantly reduced in MSA patients. Latency showed no significant difference (p = 0.083). Error correction rate (p = 0.020) was significantly lower in MSA than HCs with higher error correction reaction time (ms) (p = 0.003).

No significant differences in oculomotor parameters were observed between MSA-C and MSA-P subtypes (Supplementary Table 1).

Diagnostic values of oculomotor parameters

ROC curves were generated to investigate whether oculomotor parameters could distinguish MSA patients from HCs. The AS average velocity showed an AUC of 0.793 (sensitivity, 0.722 and specificity, 0.787), followed by SP number of deviations with an AUC of 0.772 (sensitivity, 0.794 and specificity, 0.673) and PS average velocity with an AUC of 0.732 (sensitivity, 0.602 and specificity, 0.816). A combined ROC analysis of AS average velocity, SP number of deviations and PS average velocity improved discrimination, achieving an AUC of 0.814 (sensitivity, 0.772 and specificity, 0.800) (Figure 2).

Figure 2.

ROC curves to evaluate the use of oculomotor parameters for the discrimination of MSA patients from HCs. ROC, receiver operating characteristic curve, MSA: multiple system atrophy, HC: health control, SP: smooth pursuit, PS: prosaccade, AS: antisaccade, AVG: average.

Associations between oculomotor parameters and clinical variables

Correlations between clinical characteristics and oculomotor parameters were analyzed for 100 MSA patients (Table 1&2). SP number of deviations and total deviation were positively correlated with disease duration (r = 0.331, p < 0.001 and r = 0.334, p < 0.001, respectively). PS average velocity and SP number of deviations were significantly associated with H&Y stage (r = -0.332, p < 0.001 and r = 0.325, p = 0.001, respectively). PS average velocity was negatively correlated with UMSARS total and UMSARS-II scores (r = -0.354, p < 0.001 and r = -0.379, p < 0.001, respectively). AS average velocity was negatively correlated with UMSARS-II scores (r = -0.286, p = 0.005), while SP number of deviations was positively correlated with UMSARS-IV scores (r = 0.349, p < 0.001). AS error correction rate was negatively correlated with UMSARS total and UMSARS-II scores (r = -0.413, p < 0.001 and r = −0.429, p < 0.001, respectively). Additionally, after adjusting for educational level, oculomotor parameters showed significant correlations with cognitive scores. AS accuracy and error correction rate were positively correlated with MOCA scores (r* = 0.375, p* = 0.001 and r* = 0.409, p* < 0.001, respectively).PS latency was negatively correlated with MMSE scores (r* = -0.403, p* < 0.001). PS average reaction time was correlated with MOCA scores (r* = -0.325, p* = 0.006) and FAB scores (r* = -0.340, p* = 0.002). AS error correction reaction time was associated with FAB scores (r* = -0.270, p* = 0.031) (Figure 3).

Figure 3.

Correlation analysis between oculomotor parameters and clinical variables. r values and p values calculated using Spearman's correlation coefficient analysis are indicated on each dot plot, *: adjusted for education years. PS: prosaccade, AS: antisaccade, SP: smooth pursuit, AVG: average, UMSARS: unified multiple system atrophy rating scale, MMSE: mini-mental state examination, FAB: frontal assessment battery, MoCA: Montreal cognitive assessment.

Associations between oculomotor parameters and brain atrophy

Correlation analyses were conducted in 100 MSA patients and 50 HCs with high-resolution MRI and concurrent eye-tracking data to examine relationships between regional gray matter atrophy and oculomotor performance. MSA patients exhibited significant cerebellar and frontal lobe atrophy compared to HCs (Supplementary Table 2). In multivariate analysis adjusted for age, gender, and TIV, vermis atrophy was significantly associated with PS average velocity in MSA patients. Bilateral cerebellar hemisphere volumes were positively correlated with AS average velocity. In the SP task, number of deviations was significantly correlated with atrophy in the right cerebellum, right middle frontal gyrus, and left superior frontal gyrus (p < 0.05, family-wise error correction) (Figure 4 and Table 3). No significant correlations were observed in HCs (p > 0.05, family-wise error correction). These findings indicated that oculomotor measures (PS and AS average velocity, SP number of deviations) and GMV in specific brain regions (cerebellum and frontal lobe) not only showed significant alterations in MSA but also demonstrated oculomotor-volumetric couplings compared to HCs.

Figure 4.

MRI T1W1 whole-brain voxel-based analysis VBM results. Regions of gray matter volumes associated with oculomotor parameters in MSA (p < 0.05, family-wise error correction). a: PS-Velocity_AVG (+); b: AS-Velocity_AVG (+); c: SP-Number of deviations (-). PS: prosaccade, AS: antisaccade, SP: smooth pursuit, AVG: average. The color bar represents t-values.

Table 3.

MRI T1W1 whole-brain voxel-based analysis VBM results: regions of gray matter volumes associated with oculomotor parameters in MSA (p < 0.05, family-wise error correction).

Eye movement tasks	Oculomotor parameters/Anatomic region	Cluster sizes (voxels)	Peak t value	Peak MNI Coordinates
Eye movement tasks	Oculomotor parameters/Anatomic region	Cluster sizes (voxels)	Peak t value	X	Y	Z
Prosaccade	Velocity_AVG (°/s)
	Vermis_4_5	84004	5.39	0.25	−52.25	−16.25
Antisaccade	Velocity_AVG (°/s)
	Cerebellum_Crus1_R	26643	−4.63	19.25	−77.25	−33.25
	Cerebellum_8_L	10,347	−3.88	−19.75	−71.25	−50.25
Smooth Pursuit	Number of deviation (times)
	Cerebellum_Crus1_R	4450	−4.46	38.25	−52.25	−36.25
	Frontal_Mid_2_R	11,069	−4.91	36.25	45.75	25.75
	Frontal_Sup_2_L	7149	−4.89	−15.75	40.75	39.75

Note. MNI: Montreal Neurological Institute, L: left, R: right.

Discussions

In this study, we investigated oculomotor function in patients with MSA and its relationships with clinical characteristics and regional brain volumes. Using a portable and efficient eye-tracking system integrated with VR, we found that MSA patients exhibited significantly impaired performance compared to HCs across four oculomotor tasks: fixation stability, SP, PS, and AS. Notably, AS average velocity, SP number of deviations, and PS average velocity could help differentiate MSA from HC. Correlation analyses revealed significant associations between oculomotor parameters and clinical measures. MRI-based volumetric analyses further revealed significant correlations between oculomotor parameters and regional brain atrophy in MSA patients. Specifically, PS average velocity correlated with vermis atrophy, while AS average velocity showed associations with bilateral cerebellar hemisphere volumes. Additionally, SP number of deviations was linked to atrophy in both cerebellar and frontal gyrus regions. These findings suggest that quantitative assessments of eye movement may serve as potential biomarkers for MSA diagnosis and progression monitoring.

Eye-tracking technology has emerged as a promising tool for neurological assessments, enabling precise quantification of eye movements.^11,30 Our study found that MSA patients had significant impaired SP and gaze-holding function, consistent with prior research.³¹ Specifically, MSA patients exhibited decreased SP gain, increased gaze-evoked nystagmus and frequent square-wave jerks during gaze-holding test.³² These deficits are thought to be cerebellum-related ocular signs, caused by lesions in the cerebellar flocculus and paraflocculus. Additionally, MSA patients displayed prolonged PS latencies compared to HCs. Saccadic movement impairment is common in MSA patients.¹² Prolonged reflexive saccade latency may result from lesions in brainstem burst neurons and their connections to the superior colliculus, delaying saccade initiationn.^33,34 Although slowed saccades were once considered exclusionary for MSA,³⁵ recent studies confirm prolonged saccades and reduced saccadic velocity in MSA compared to HCs.³⁶ The cerebellum also plays a role in the function of the cortical-superior colliculus circuit by connecting with brainstem saccade generators to modulate the amplitude and peak velocity of saccades.³⁷

The AS task is a well-established paradigm for assessing frontostriatal function by measuring the ability to suppress reflexive saccades toward a sudden peripheral target and generate voluntary saccade in the opposite direction.³⁸ This task evaluates both inhibitory control and voluntary saccade generation.³⁹ While AS impairments, such as increased error rates and prolonged latencies, are well-documented in Parkinson's disease (PD) and Alzheimer's disease (AD),^20,40,41 AS performance in MSA remains understudied. Our findings revealed that MSA patients exhibited decreased AS accuracy and error correction rate, longer reaction times and error correction times than HCs, indicating impairment of the frontalstriatal network, a feature shared across these neurodegenerative disorders. Compared to PD, MSA patients may display more severe AS dysfunction, likely due to combined degeneration of fronto-basal ganglia and cerebellar-brainstem pathways, resulting in pronounced deficits in inhibitory control and voluntary saccade generation.³⁸

Given the pronounced oculomotor deficits in MSA patients, we further evaluated their potential as diagnostic biomarkers. Reliable biomarkers are important for MSA diagnosis, as it presents complex and variable symptoms.⁴² Compared to clinical scales, quantitative oculomotor parameters provide sensitive and reliable measures.^43,44 Our study revealed that AS and PS average velocity, SP number of deviations effectively differentiated MSA patients from HCs. Oculomotor abnormalities have been reported in idiopathic rapid eye movement sleep behavior disorder (iRBD),^45,46 a prodromal stage of α-synucleinopathies, suggesting their potential as early neurodegenerative markers. However, oculomotor parameters showed limited ability to differentiate MSA-P from MSA-C subtypes, aligning with previous studies that reported no significant differences across multiple oculomotor parameters.^47–49 This may reflect a shared underlying pathophysiology, as postmortem studies have demonstrated that olivopontocerebellar neurons are affected in both subtypes even at early stages, establishing cerebellar dysfunction as a common pathological substrate.^50,51 Although one study reported differences,¹² we speculate that this discrepancy may stem from variations in patient cohorts or experimental conditions.

The relationship between eye movements and clinical scales in MSA has been scarcely reported. Our study found that SP, PS and AS ocular parameters were strongly correlated with clinical scores. These findings suggest that progressive alterations in oculomotor neurocircuitries are linked to MSA pathogenesis and neurodegeneration.⁵² Beyond motor burden,⁵³ eye movement alterations also reflect cognitive impairments. Stuart et al. demonstrated that saccadic characteristics help identify patients who might be at risk for rapid cognitive decline.⁵⁴ Our study identified associations between PS and AS performance and cognitive assessments, indicating that saccade performance profiles may reflect cognitive function in MSA patients. As potential treatments to slow MSA progression are developed,⁵⁵ our findings suggest that portable VR-based quantitative ocular measures could serve as valuable tools to assess disease progression and treatment effects in future trials.

Neuroimaging analysis revealed significant GM loss in the cerebellum and frontal lobe in MSA patients compared to HCs, consistent with prior studies.^56,57 However, previous neuroimaging studies only reported correlations between SP and cerebellar/ pontine volumes.^21,22 Here, we expanded these findings by revealing associations between multiple oculomotor parameters and regional atrophy. Specifically, SP performance correlated with atrophy in both cerebellum and frontal cortex, indicating that SP impairment in MSA involves both motor control and higher-order cognitive regulation.⁵⁸ Reduced PS average velocity was linked to atrophy in the cerebellar vermis, a region that receives input from the nucleus reticularis tegmenti pontis and modulates brainstem burst neurons via the fastigial oculomotor region, thereby disrupting saccadic regulation.⁵⁹ This is distinct from PD, where oculomotor deficits primarily reflect dysfunction within the basal ganglia-thalamocortical circuits.²⁰ Similarly, bilateral cerebellar hemisphere volumes positively correlated with AS velocity, aligning with evidence that cerebellar hemispheres support higher-order oculomotor control.^60–62 Overall, these findings highlight the importance of the “cerebello-prefrontal circuit” in MSA-related oculomotor deficits and reflect underlying pathological changes. Thus, oculomotor parameters may serve as potential biomarkers for detecting circuit dysfunction and monitoring the efficacy of interventions in future clinical trials.

The VR-based eye-tracking assessment offers several translational and clinical values. First, it enabled a novel finding: markedly reduced PS and AS average velocities, combined with increased SP number of deviations in MSA. These parameters demonstrated diagnostic value and were significantly associated with clinical severity scores. Unlike traditional eye-tracking systems, the VR headset allows for the presentation of visual motion stimuli in a head-unrestrained, immersive environment, which may reveal oculomotor impairments through more naturalistic saccades and enhance the ecological validity of these parameters as potential biomarkers.^18,63 Second, this approach provides an objective, portable, and user-friendly alternative, enabling efficient testing in clinical or outpatient settings with minimal equipment. This enhances accessibility for patients with mobility issues or in resource-limited environments and holds promise for future remote telemedicine applications, as studies have shown that eye-tracking VR interfaces can emulate physician tasks for remote diagnosis of neurodegenerative diseases.⁶⁴ Finally, our findings reveal specific brain-behavior associations, demonstrating that oculomotor parameters are closely linked to atrophy in distinct brain regions, particularly the cerebellum. To our knowledge, this is the first study to use VR eye-tracking to explore MSA oculomotor dysfunction and its relationships with regional brain volumes, positioning these parameters as novel, non-invasive proxies for assessing the functional integrity of specific neural circuits and facilitating rapid evaluation for early diagnosis and progression monitoring in MSA.

The current study has several limitations. First, the cross-sectional design limits conclusions about the progression of oculomotor dysfunction and gray matter atrophy in MSA over time. Future longitudinal studies are needed to track oculomotor and imaging changes in MSA cohorts. Second, our control group did not include patients with other neurodegenerative disorders such as PD and progressive supranuclear palsy (PSP), which limits the validation of the diagnostic specificity of the oculomotor parameters in differentiating MSA from other neurodegenerative conditions. Future studies involving additional disease control groups are needed to further explore the differential diagnostic value of these oculomotor parameters.

In conclusion, our study provides comprehensive insights into the VR-based quantitative oculomotor analysis and its association with regional brain atrophy in MSA. Notably, parameters such as PS and AS average velocities, and SP number of deviations, hold promise as potential biomarkers for MSA diagnosis and disease severity assessment.

Supplemental Material

sj-docx-1-pkn-10.1177_1877718X251412237 - Supplemental material for VR-based quantitative oculomotor analysis and association with regional brain atrophy in MSA

Supplemental material, sj-docx-1-pkn-10.1177_1877718X251412237 for VR-based quantitative oculomotor analysis and association with regional brain atrophy in MSA by You Fu, Linlin Wan, Zhao Chen, Chunrong Wang, Daji Chen, Riwei Ouyang, Xiaokang Wu, Xiafei Long, Kefang Du, Xinying Xiao, Ruqing He, Linliu Peng, Rong Qiu, Beisha Tang and Hong Jiang in Journal of Parkinson's Disease

Footnotes

Acknowledgements

We thank all the participants for their involvement in this study.

Authors’ roles

YF and HJ: Design; YF, LLW, ZC, CRW, DJC, RWO, XKW, XFL, KFD, XYX and RQH: Execution and Analysis; YF: Writing of the first draft; LLW, LLP, RQ, BST and HJ: Editing of final version of the manuscript.

Financial disclosure/funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the National Key R&D Program of China (No. 2021YFA0805200 to H Jiang), the National Natural Science Foundation of China (U25A2058, 82571433, 82171254 to H Jiang; 82571632, 82371272 to Z Chen; 82301628 to L Wan; 82401496 to L Peng), the Natural Science Foundation of Hunan Province (No. 2024JJ3050 to H Jiang; 2025JJ20078 to Z Chen; 2024JJ6638 to L Wan), the Major Scientific Research Project for High-level Health Talent in Hunan Province (No. R2023047 to H Jiang), Furong Lab Research Project (No. 2023SK2084 to H Jiang), MIIT Project for Open Competition and Leadership in AI Medical Device Innovation (No. 2025-AISJK-01009 to H Jiang), the Central South University Research Programme of Advanced Interdisciplinary Study (No. 2023QYJC010 to H Jiang), Postdoctoral Fellowship Program of CPSF (GZB20230870 to L Peng) and China Postdoctoral Science Foundation (2024M753690 to L Peng).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Supplemental material

Supplemental material for this article is available online.

References

Poewe

Stankovic

Halliday

, et al. Multiple system atrophy. Nat Rev Dis Primers 2022; 8: 56.

Fanciulli

Wenning

. Multiple-system atrophy. N Engl J Med 2015; 372: 249–263.

Trojanowski

Revesz

. Proposed neuropathological criteria for the post mortem diagnosis of multiple system atrophy. Neuropathol Appl Neurobiol 2007; 33: 615–620.

Krismer

Wenning

. Multiple system atrophy: insights into a rare and debilitating movement disorder. Nature Reviews Neurology 2017; 13: 232–243.

Stankovic

Quinn

Vignatelli

, et al. A critique of the second consensus criteria for multiple system atrophy. Mov Disord 2019; 34: 975–984.

Foubert-Samier

Pavy-Le Traon

Guillet

, et al. Disease progression and prognostic factors in multiple system atrophy: a prospective cohort study. Neurobiol Dis 2020; 139: 104813.

Zhang

Hou

Cao

, et al. Longitudinal evolution of motor and non-motor symptoms in early-stage multiple system atrophy: a 2-year prospective cohort study. BMC Med 2022; 20: 46.

Krismer

Seppi

Jönsson

, et al. Sensitivity to change and patient-centricity of the unified multiple system atrophy rating scale items: a data-driven analysis. Mov Disord 2022; 37: 1425–1431.

Koga

Dickson

. Recent advances in neuropathology, biomarkers and therapeutic approach of multiple system atrophy. J Neurol Neurosurg Psychiatry 2018; 89: 175–184.

10.

Riek

Visanji

Pitigoi

, et al. Multimodal oculomotor assessment reveals prodromal markers of Parkinson's disease in non-manifesting LRRK2 G2019S mutation carriers. NPJ Parkinsons Dis 2024; 10: 34.

11.

Anderson

MacAskill

. Eye movements in patients with neurodegenerative disorders. Nature Reviews Neurology 2013; 9: 74–85.

12.

Zhou

Sun

Wei

, et al. Quantitative assessment of oculomotor function by videonystagmography in multiple system atrophy. Clin Neurophysiol 2022; 141: 15–23.

13.

Coe

Munoz

. Mechanisms of saccade suppression revealed in the anti-saccade task. Philos Trans R Soc Lond B Biol Sci 2017; 372: 20160192.

14.

Munoz

Everling

. Look away: the anti-saccade task and the voluntary control of eye movement. Nat Rev Neurosci 2004; 5: 218–228.

15.

Waldthaler

Tsitsi

Svenningsson

. Vertical saccades and antisaccades: complementary markers for motor and cognitive impairment in Parkinson's disease. npj Parkinson's Disease 2019; 5: 11.

16.

Barbosa

Kaski

Castro

, et al. Saccadic direction errors are associated with impulsive compulsive behaviours in Parkinson's disease patients. Journal of Parkinson’s Disease 2019; 9: 625–630.

17.

Zhang

Pan

, et al. A portable and efficient dementia screening tool using eye tracking machine learning and virtual reality. npj Digital Medicine 2024; 7: 19.

18.

Adhanom

MacNeilage

Folmer

. Eye tracking in virtual reality: a broad review of applications and challenges. Virtual Real 2023; 27: 1481–1505.

19.

Imaoka

Flury

de Bruin

. Assessing saccadic eye movements with head-mounted display virtual reality technology. Front Psychiatry 2020; 11: 572938.

20.

Antoniades

Spering

. Eye movements in Parkinson's disease: from neurophysiological mechanisms to diagnostic tools. Trends Neurosci 2024; 47: 71–83.

21.

Vintonyak

Gorges

Müller

, et al. Patterns of eye movement impairment correlate with regional brain atrophy in neurodegenerative parkinsonism. Neurodegener Dis 2017; 17: 117–126.

22.

Gorges

Maier

Rosskopf

, et al. Regional microstructural damage and patterns of eye movement impairment: a DTI and video-oculography study in neurodegenerative parkinsonian syndromes. J Neurol 2017; 264: 1919–1928.

23.

Wenning

Stankovic

Vignatelli

, et al. The movement disorder society criteria for the diagnosis of multiple system atrophy. Mov Disord 2022; 37: 1131–1148.

24.

Wenning

Tison

Seppi

, et al. Development and validation of the unified multiple system atrophy rating scale (UMSARS). Mov Disord 2004; 19: 1391–1402.

25.

Hoops

Nazem

Siderowf

, et al. Validity of the MoCA and MMSE in the detection of MCI and dementia in Parkinson disease. Neurology 2009; 73: 1738–1745.

26.

Aiello

Esposito

Gramegna

, et al. The frontal assessment battery (FAB) and its sub-scales: validation and updated normative data in an Italian population sample. Neurol Sci 2022; 43: 979–984.

27.

Chen

, et al. Artificial intelligence-assisted oculo-gait measurements for cognitive impairment in cerebral small vessel disease. Alzheimers Dement 2024; 20: 8516–8526.

28.

Lin

Yang

, et al. A detection model of cognitive impairment via the integrated gait and eye movement analysis from a large Chinese community cohort. Alzheimers Dement 2024; 20: 1089–1101.

29.

Whitwell

. Voxel-based morphometry: an automated technique for assessing structural changes in the brain. J Neurosci 2009; 29: 9661–9664.

30.

Gorges

Pinkhardt

Kassubek

. Alterations of eye movement control in neurodegenerative movement disorders. J Ophthalmol 2014; 2014: 658243.

31.

Zhou

Huang

, et al. Clinical value of video oculomotor evaluation in the differential diagnosis of multiple system atrophy and Parkinson's disease. Brain Behav 2024; 14: e3510.

32.

Zhou

Wang

, et al. The differential diagnostic value of a battery of oculomotor evaluation in Parkinson's disease and multiple system atrophy. Brain Behav 2021; 11: e02184.

33.

Zhou

Wei

Jiang

, et al. Abnormal ocular movement in the early stage of multiple-system atrophy with predominant parkinsonism distinct from Parkinson's disease. Journal of Clinical Neurology 2024; 20: 37–45.

34.

May

. The mammalian superior colliculus: laminar structure and connections. Prog Brain Res 2006; 151: 321–378.

35.

Anderson

Luxon

Quinn

, et al. Oculomotor function in multiple system atrophy: clinical and laboratory features in 30 patients. Mov Disord 2008; 23: 977–984.

36.

Wang

, et al. Smooth pursuit and reflexive saccade in discriminating multiple-system atrophy with predominant parkinsonism from Parkinson's disease. Journal of Clinical Neurology 2024; 20: 194–200.

37.

Terao

Tokushige

Inomata-Terada

, et al. Differentiating early Parkinson's disease and multiple system atrophy with parkinsonism by saccade velocity profiles. Clin Neurophysiol 2019; 130: 2203–2215.

38.

Brooks

Klier

Red

, et al. Slowed prosaccades and increased antisaccade errors as a potential behavioral biomarker of multiple system atrophy. Front Neurol 2017; 8: 20170620.

39.

Riek

Brien

Coe

, et al. Cognitive correlates of antisaccade behaviour across multiple neurodegenerative diseases. Brain Communications 2023; 5: fcad049.

40.

Levy N

Lavidor

Vakil

. Prosaccade and antisaccade paradigms in persons with Alzheimer's disease: a meta-analytic review. Neuropsychol Rev 2018; 28: 16–31.

41.

Gallea

Wicki

Ewenczyk

, et al. Antisaccade, a predictive marker for freezing of gait in Parkinson's disease and gait/gaze network connectivity. Brain 2021; 144: 504–514.

42.

Palma

Norcliffe-Kaufmann

Kaufmann

. Diagnosis of multiple system atrophy. Auton Neurosci 2018; 211: 15–25.

43.

Garces

Antoniades

Sobanska

, et al. Quantitative oculomotor assessment in hereditary ataxia: systematic review and consensus by the ataxia global initiative working group on digital-motor biomarkers. Cerebellum 2024; 23: 896–911.

44.

Wunderlich

Behler

Dreyhaupt

, et al. Diagnostic value of video-oculography in progressive supranuclear palsy: a controlled study in 100 patients. J Neurol 2021; 268: 3467–3475.

45.

Iranzo

Santamaria

Tolosa

. Idiopathic rapid eye movement sleep behaviour disorder: diagnosis, management, and the need for neuroprotective interventions. Lancet Neurol 2016; 15: 405–419.

46.

Chen

Zhou

Jiang

, et al. Impaired ocular tracking and cortical atrophy in IdiopathicRapid eye MovementSleep behavior disorder. Mov Disord 2022; 37: 972–982.

47.

Terao

Fukuda

Tokushige

, et al. Is multiple system atrophy with cerebellar ataxia (MSA-C) like spinocerebellar ataxia and multiple system atrophy with parkinsonism (MSA-P) like Parkinson's disease? - A saccade study on pathophysiology. Clin Neurophysiol 2016; 127: 1491–1502.

48.

Pinkhardt

Kassubek

Süssmuth

, et al. Comparison of smooth pursuit eye movement deficits in multiple system atrophy and Parkinson's disease. J Neurol 2009; 256: 1438–1446.

49.

Isozaki

Tobisawa

Naito

, et al. Eye movement disturbance in multiple system atrophy: chronological study of 50 patients. Rinsho Shinkeigaku 2012; 52: 218–226.

50.

Jellinger

. Neuropathology of multiple system atrophy: new thoughts about pathogenesis. Mov Disord 2014; 29: 1720–1741.

51.

Ozawa

Paviour

Quinn

, et al. The spectrum of pathological involvement of the striatonigral and olivopontocerebellar systems in multiple system atrophy: clinicopathological correlations. Brain 2004; 127: 2657–2671.

52.

Chen

Feng

, et al. Oculomotor deficits in spinocerebellar ataxia type 3: potential biomarkers of preclinical detection and disease progression. CNS Neurosci Ther 2017; 23: 321–328.

53.

Waldthaler

Stock

Student

, et al. Antisaccades in Parkinson's disease: a meta-analysis. Neuropsychol Rev 2021; 31: 628–642.

54.

Stuart

Lawson

Yarnall

, et al. Pro-saccades predict cognitive decline in Parkinson's disease: ICICLE-PD. Mov Disord 2019; 34: 1690–1698.

55.

Sivalingam

. Advances in understanding biomarkers and treating neurological diseases - role of the cerebellar dysfunction and emerging therapies. Ageing Res Rev 2024; 101: 20240927.

56.

Brenneis

Egger

Scherfler

, et al. Progression of brain atrophy in multiple system atrophy. J Neurol 2007; 254: 191–196.

57.

Krismer

Péran

Beliveau

, et al. Progressive brain atrophy in multiple system atrophy: a longitudinal, multicenter, magnetic resonance imaging study. Mov Disord 2024; 39: 119–129.

58.

Fukushima

Barnes

. Clinical application of eye movement tasks as an aid to understanding Parkinson's disease pathophysiology. Exp Brain Res 2017; 235: 1309–1321.

59.

Jensen

Beylergil

Shaikh

. Slow saccades in cerebellar disease. Cerebellum Ataxias 2019; 6: 20190117.

60.

Kunimatsu

Suzuki

Tanaka

. Implications of lateral cerebellum in proactive control of saccades. J Neurosci 2016; 36: 7066–7074.

61.

Peterburs

Thürling

Rustemeier

, et al. A cerebellar role in performance monitoring - evidence from EEG and voxel-based morphometry in patients with cerebellar degenerative disease. Neuropsychologia 2015; 68: 139–147.

62.

Fielding

Corben

Cremer

, et al. Disruption to higher order processes in Friedreich ataxia. Neuropsychologia 2010; 48: 235–242.

63.

Merzon

Pettersson

Aronen

, et al. Eye movement behavior in a real-world virtual reality task reveals ADHD in children. Sci Rep 2022; 12: 20221124.

64.

Orlosky

Itoh

Ranchet

, et al. Emulation of physician tasks in eye-tracked virtual reality for remote diagnosis of neurodegenerative disease. IEEE Trans Vis Comput Graph 2017; 23: 1302–1311.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB