Correlations between SSQ Scores and ECG Data during Virtual Reality Walking by Display Type

Abstract

To encourage the application of virtual reality (VR) in physical rehabilitation, this study analyzed the occurrence of motion sickness when walking on a treadmill in virtual straight paths presented on two types of displays (screen and head-mounted displays (HMDs)) at a constant speed of 3.6 km/h. The simulator sickness questionnaire (SSQ) scores, which indicate motion sickness, were collected from the participants. In addition, the heart rate (HR) and heart rate variability (HRV; RMSSD and LF/HF ratio) were measured from electrocardiogram data. The correlations between the SSQ scores and HR and HRV were examined to identify a reliable variable for evaluating motion sickness. The SSQ scores were used to classify the data into the motion-sickness and no-motion-sickness groups. The data were classified into the motion-sickness group if a minimum difference of 15 points existed between the walking and baseline phases when using the screen and HMD; otherwise, the data were classified into the no-motion-sickness group. The HR and LF/HF ratio were higher, whereas the RMSSD was lower in the motion-sickness group. Moreover, within the motion-sickness group, the reduction in RMSSD and increase in HR and LF/HF ratio were greater with the HMD than with the screen. Regression analysis was performed on the HR, HRV, and SSQ scores to differentiate between the motion-sickness and no-motion-sickness groups. The regression analysis results showed a high negative correlation between the SSQ score and RMSSD. The results of this study can assist in controlling the occurrence of motion sickness in VR-based applications.

1. Introduction

The demand for head-mounted displays (HMDs) for viewing virtual reality (VR) content has been increasing annually. In particular, VR technology has seen significant growth in the wake of COVID-19, owing to remote meetings becoming increasingly common. Moreover, the increase in the number of people who want to gain various experiences from the comfort of their homes through VR devices has led to a rapid increase in the number of VR users [1].

However, several side effects of VR have been reported, the most common being cybersickness, also known as simulator sickness or motion sickness, which occurs when using VR; it is accompanied by symptoms similar to those of actual motion sickness, such as eye fatigue, disorientation, and nausea [2]. Cybersickness is a condition characterized by symptoms such as nausea, dizziness, and headaches experienced by users in virtual reality (VR) environments, with the sensory conflict theory being reported as the primary cause. This refers to the discrepancy between visual information in the virtual environment and the motion and position senses from the inner ear, leading to motion sickness. Additionally, factors in VR environments such as parabolic motion representation, update latency, limited field of view, and individual differences among users have also been reported as causes of cybersickness [3]. Although studies have been conducted to identify and quantify the causes of cybersickness, its symptoms and degree vary among individuals, making research difficult.

In most previous studies, a moving virtual environment was presented to stationary seated subjects, and the degree of motion sickness was measured. For subjective evaluations, the VR sickness questionnaire, which includes the simulator sickness questionnaire (SSQ), was used [4], and the motion sickness susceptibility questionnaire (MSSQ) was used to determine the individual differences in motion sickness due to various stimuli [5]. For objective evaluations, heart rate (HR) and heart rate variability (HRV) analyses through electrocardiograms (ECGs) were primarily used. In addition, electroencephalographs, which indirectly measure the electrical activity in the neurons of the brain [6,7], breathing rate [8], and body temperature [9], were used for evaluating motion sickness.

One of these biosignals, the HRV, indicates the changes in the HR intervals related to autonomic nervous system control, including the sympathetic and parasympathetic nervous systems [10]. Parasympathetic activity slows down the HR; conversely, sympathetic activity increases the HR. The dynamic balance between the parasympathetic and sympathetic activities causes continuous oscillation of the HR, which is called the HRV [11]. Physiological indicators such as the HRV parameters allow researchers to quickly ascertain the physical conditions of the study participants. HRV parameters can be used to examine the condition of a study participant based on their parasympathetic and sympathetic activities; therefore, the HRV parameters that are most highly correlated with cybersickness should be studied.

Accordingly, questionnaires and various biosignal measurements are being used to examine these correlations with motion sickness; however, the majority of the existing cybersickness studies have been performed in a state of limited movement, and studies on motion sickness during regular movements such as walking are limited. In particular, the use of VR in walking rehabilitation is currently increasing; consequently, it has become necessary to examine cybersickness and control the factors affecting it [12,13]. Parijat et al. (2011) examined the gait variables in 16 healthy seniors after walking on a treadmill while wearing or not wearing an HMD displaying a walking trail environment [12]. Increased variability in the gait variable and joint flexion were observed within 10 min of performing VR walking when compared with normal walking; however, after 15 min, the walking characteristics became similar to those of normal walking. This confirms that the acclimation time is important when designing VR-based training. In the case of Lheureux et al. (2020), 10 Parkinson’s patients in their 60s performed treadmill walking with track walking/treadmill walking/VR, and we tried to compare the walking variables at that time [13]. As a result, VR walking has no difference in stride time variability (i.e., coefficient of variation, CV) or fractal dynamics (i.e., α value of detailed function analysis, DFA) under different walking conditions, but the step length and gait cadence were improved compared to treadmill walking. These studies show that VR can be used to implement gait rehabilitation. When VR is used to perform rehabilitation, side effects such as cybersickness are considered a major limitation. To use VR in various fields in addition to rehabilitation, there is a need for research on cybersickness during VR use that is accompanied by movement, namely, walking. Many studies have been conducted on cybersickness that occurs when viewing virtual environments; however, research on cybersickness that occurs when viewing virtual environments accompanied by regular movements is lacking. Therefore, this study aimed to use a subjective evaluation (SSQ) and an objective evaluation (ECG) to determine the cybersickness that occurs when subjects are presented with a virtual straight path on two types of displays (screen and HMD) and walk at a constant speed on a treadmill. In addition, this study analyzed the difference in HR and HRV (root mean square of the successive differences (RMSSD), LF/HF ratio) between the motion-sickness group and no-motion-sickness group, as classified using the SSQ, and analyzed the correlation between the SSQ score that determined the presence of motion sickness and HR and HRV to find reliable variables that allow motion sickness to be evaluated via ECG.

2. Research Methods

2.1. Experimental Design

Eleven men in their 20s (age 23.7 ± 2.5, height 176.5 ± 5.5 cm, weight 74.6 ± 12 kg) participated in this study. The protocol for the research project was approved by the Institutional Review Committee of Konkuk University, where this study was performed, and conforms to the provisions of the Declaration of Helsinki (IRB number: 7001355-202307-HR-675).

In the experiments, Unity 3D (Unity Technologies, San Francisco, CA, USA) was used to present the subjects with a virtual straight walking path (Figure 1), and they walked continuously on a treadmill (DRAX, RX9200, Anyang, Gyeonggi-do, Republic of Korea). The virtual scene was presented to the subjects via a screen (160 cm × 120 cm) and an HMD (Oculus Quest 2, Meta, weight: 503 g) (Figure 2).

The experiment was conducted under two conditions: screen walking, where participants walked on a treadmill while looking at a screen (160 cm × 120 cm), and VR walking, where participants walked on a treadmill while wearing a Head-Mounted Display (HMD). In both conditions, the screen moved in sync with the participant’s walking speed on the treadmill. The experimental design for both conditions was identical, as shown in Figure 3. The baseline segment involved participants standing still and looking at the screen for 1 min in a rest state. During the walking segment, participants performed a total of 15 min of walking on a straight pathway, with the walking speed fixed at 3.6 km/h. The rest segment was a break period, during which participants were instructed to stand still and look at the screen to return to a rested state. The experiment was conducted randomly, with a 30 min rest period after the walking experiment of one condition before performing the walking experiment of the other condition. Each condition was performed once, without repetition.

2.2. Data Measurement and Analysis

The subjective and objective indicators of cybersickness were measured under two conditions (screen-assisted walking and HMD-assisted walking). The Simulator Sickness Questionnaire (SSQ) survey was used as a subjective indicator, and the most commonly used evaluation method to measure subjective motion sickness in cybersickness [14], and there are many variations in the type or number of items. Therefore, although the total score can vary in each study, the SSQ score is considered highly reliable because it includes evaluation items that consider various circumstances leading to cybersickness, and it is not only affected by the visualization screen but also by control methods such as joysticks [15]. The SSQ consists of five main items, namely, eyestrain, overall discomfort, nausea, concentration, and headache, as well as 28 sub-items that belong to these items; a score of 1–7 is entered for each item. In this study, the SSQ questionnaires were completed by each participant directly after the baseline and walking phases in each of the two experimental conditions (screen and HMD). The SSQ test for each phase took about 3 min.

ECGs were measured as the objective indicators, and the Trigno EKG sensor (Delsys, Natick, MA, USA) was used as the measurement equipment. For the experiment, two electrodes were used, with one attached under each side of the chest. The HR and HRV were extracted using HRV analysis (ANS Lab Tools). For the HRV, the RMSSD and LF/HF ratio were used to examine the activation of the sympathetic and parasympathetic nerves. The RMSSD is the root mean square of the differences in the RR interval and is calculated by using the time difference between consecutive heartbeats in milliseconds, squaring each value, finding the average of the result, and determining the square root of the total. It reflects the beat-to-beat variance in the HR and is used as a domain measurement to estimate the changes in the HRV [16]. A high value of the LF/HF ratio indicates that the sympathetic nerves are activated or that the activation of the parasympathetic nerves is blocked. In short, this value can help quantify the overall balance between the sympathetic and parasympathetic nervous systems. The HR, RMSSD, and LF/HF ratio measured in the experiments under the two conditions were analyzed based on variations in the values, which were determined by subtracting the baseline data from the walking interval data.

To analyze the correlation between the occurrence of cybersickness and HR and HRV, the SSQ score of each subject was used to divide the subjects into two groups (motion sickness and no motion sickness) and perform an analysis. The SSQ scores after the baseline phase and those after the walking phase were compared; subjects with a score difference of less than 15 points were classified as the no-motion-sickness group, whereas those with a score difference of more than 15 points were classified as the motion-sickness group. Thus, six subjects were classified under the no-motion-sickness group (age 23.5 ± 5.8, height 175.2 ± 5.8 cm, weight 68.2 ± 8.8 kg), and five subjects were classified under the motion-sickness group (age 24 ± 1.4, height 178.0 ± 5.3 cm, weight 82.4 ± 11.1 kg) according to the differences in their SSQ scores, and an analysis of the HR and HRV according to the occurrence of motion sickness was performed.

To examine the differences in SSQ scores, HR, RMSSD, and LF/HF ratio differences between the two conditions (screen and HMD) and the two groups (motion-sickness group, no-motion-sickness group), we conducted a two-way ANOVA using SPSS 25 (IBM, Armonk, NY, USA). Specifically, we first performed paired-sample t-tests to assess SSQ scores, HR, RMSSD, and LF/HF ratio differences within each group across the two conditions. Second, we utilized independent-sample t-tests to examine SSQ scores, HR, RMSSD, and LF/HF ratio differences between the two groups within each condition. Lastly, to determine whether objective indicators such as HR, RMSSD, and LF/HF ratio differences could also be used to distinguish between the motion-sickness and no-motion-sickness groups, in addition to the subjective indicator of SSQ scores, we conducted linear regression analysis. This analysis aimed to clarify whether higher SSQ scores correspond to higher or lower values of HR, RMSSD, or LF/HF ratio. Using SSQ scores as the dependent variable and the remaining variables (HR, RMSSD, LF/HF ratio differences) as independent variables, we sought to identify which variable best represents SSQ scores and could therefore serve as an objective indicator for distinguishing between the motion-sickness and no-motion-sickness groups.

3. Experimental Results

Figure 4 presents a graph of the SSQ scores of the two groups under the two experimental conditions (screen-assisted walking and HMD-assisted walking). The score of the no-motion-sickness group was 2.3 for screen-assisted walking and 4.2 for HMD-assisted walking, that is, the score with the HMD was higher by 1.9. In the motion-sickness group, the score was 21 with the screen and 34.6 with the HMD, that is, an increase of 13.6. Statistical analysis confirmed that the increase in the SSQ scores between the two groups was significant (p < 0.001). The t-test showed significant results under the two conditions for the motion-sickness group (p = 0.005), and there was a significant difference in the SSQ score between the two groups during the HMD-assisted walking phase (p < 0.001).

Table 1. HR, RMSSD, and LF/HF ratio for motion-sickness group and no-motion-sickness group for each phase (baseline, walking), and variations between the phases during virtual environment walking under two conditions (screen and HMD).

Screen		Baseline Phase			Walking Phase			Variation (Walking Phase—Baseline Phase)
		HR	RMSSD	LF/HF	HR	RMSSD	LF/HF	HR	RMSSD	LF/HF
no-motion-sickness	#1	69	20.4	2.1	93.5	17.9	5.2	24.5	−2.5	3.1
	#2	68	12.4	1.1	77.5	8.6	3.7	9.5	−3.8	2.6
	#3	111.6	8.3	1.1	115.4	6.3	3.5	3.8	−2	2.4
	#4	66.3	30.5	5.3	83.1	31.8	1.7	16.8	1.3	−3.7
	#5	113	32.8	2.1	109.1	31.4	2.9	−3.9	−1.4	0.8
	#6	106.6	5.7	5.4	107.3	6.4	6.7	0.7	0.7	1.3
	Avg. ± S.D	89.1 ± 23.5	18.4 ± 11.5	2.9 ± 2.0	97.7 ± 15.3	17.1 ± 12.0	4.0 ± 1.8	8.58 ± 10.6	−1.3 ± 1.9	1.1 ± 2.5
motion-sickness	#1	74.9	14.2	3.5	83.9	11.8	5	9.0	−2.4	1.5
	#2	70.9	41.8	1.7	84.4	26.2	2.1	13.5	−15.6	0.4
	#3	73.4	58.6	0.4	87.2	27.2	3.5	13.8	−31.4	3.1
	#4	84.9	35.8	2.1	98.5	45.6	2.1	13.6	9.8	0
	#5	62.3	50.6	3	73.9	53.9	2.3	11.6	3.3	−0.7
	Avg. ± S.D	73.3 ± 8.1	40.2 ± 16.9	2.1 ± 1.2	85.6 ± 8.8	33.0 ± 16.8	3.0 ± 1.3	12.3 ± 2.0	−7.3 ± 16.4	0.86 ± 1.5
HMD		Baseline phase			Walking phase			Variation (Walking phase—Baseline phase)
		HR	RMSSD	LF/HF	HR	RMSSD	LF/HF	HR	RMSSD	LF/HF
no-motion-sickness	#1	75.3	23.3	2.2	84.6	15.3	2.8	9.3	−8	0.7
	#2	65.8	6.7	3.4	80.3	6.8	5.4	14.5	0.1	2
	#3	111.9	5.4	1.4	119.8	−11.5	4.5	7.9	−17	3.2
	#4	66.1	38.9	1.2	78.2	27.2	3.8	12.1	−11.7	2.6
	#5	107.2	37.5	3.7	115	49	1.5	7.8	11.6	−2.2
	#6	106.6	5.7	5.4	107.5	6	6.6	0.9	0.3	1.2
	Avg. ± S.D	88.8 ± 22.0	19.6 ± 15.9	2.9 ± 1.6	97.6 ± 18.6	15.5 ± 20.8	4.1 ± 1.8	8.7 ± 4.7	−4.1 ± 10.2	1.3 ± 1.9
motion-sickness	#1	67.1	26	2.1	82.4	10.1	6.5	15.3	−15.9	4.40
	#2	64.4	46.8	2	88.3	24.9	4	23.9	−21.9	2.00
	#3	65.1	48.3	0.6	83.9	19.6	2.1	18.8	−28.7	1.50
	#4	94.1	42.5	1.5	108.5	42.5	2.9	14.4	−0.1	1.40
	#5	70.5	35.3	2.2	84.5	20	12.4	14.0	−15.3	10.20
	Avg. ± S.D	72.2 ± 12.4	39.8 ± 9.2	1.7 ± 0.7	89.5 ± 10.8	23.4 ± 11.9	5.6 ± 4.2	17.3 ± 4.2	−16.4 ± 10.6	3.9 ± 3.7

presents the HR, RMSSD, and LF/HF ratio differences for the two groups (no motion sickness and motion sickness) under the two conditions. Figure 5a shows the HR difference between the two groups (no motion sickness and motion sickness) under the two conditions. The results of a two-way ANOVA showed that only the difference in HR was significant between the two groups (p = 0.047). That is, there was a significant increase in HR difference in the motion-sickness group when compared with that in the no-motion-sickness group. In addition, the t-test results showed that there was a significant increase in HR when the HMD was used compared to when the screen was used in the motion-sickness group (p = 0.04). Moreover, the motion-sickness group showed a significant increase in HR difference when compared with that of the no-motion-sickness group when wearing the HMD (p = 0.006).

Figure 5b shows the RMSSD data results for the two conditions (HMD and screen) and two groups. The two-way ANOVA results confirmed the difference in RMSSD between the two groups (p = 0.037). That is, the reduction in the RMSSD of the motion-sickness group when compared with that of the no-motion-sickness group was significant. The t-test results showed that the RMSSD decreased by a greater extent during walking with the HMD than during walking with the screen in the motion-sickness group (p = 0.027). In addition, the RMSSD decreased by a greater extent in the motion-sickness group than in the no-motion-sickness group when wearing the HMD (p = 0.012).

Figure 5c shows the LF/HF ratios; the analysis of the LH/HF ratios for the two groups under the two conditions did not show statistical significance for the groups or conditions. The t-test results showed that the increase in the LF/HF ratio in the motion-sickness group was greater during screen-assisted walking than during HMD-assisted walking (p = 0.04), and the increase in the LF/HF ratio was greater in the motion-sickness group than in the non-motion-sickness group when walking with the HMD (p = 0.05).

In the linear regression analysis, a stepwise method was used, and the SSQ scores were designated as dependent variables, whereas the HR, RMSSD, and LF/HF ratio were designated as independent variables. From the analysis, the RMSSD was selected as the variable of interest, and the results were significant at p < 0.001 (

Table 2. Regression analysis results for SSQ scores and difference of HR, RMSSD, and LF/HF ratio (stepwise): entered variables and excluded variables.

Entered Variables	B	t	Sig.
RMSSD	−1.045	−5.676	0.000
Excluded Variables	B	t	Sig.
HR	0.223	1.515	0.146
LF/HF	−0.93	−0.605	0.552

). These results were found to have an explanatory power of approximately 62%, with R² = 0.617 (Figure 6).

4. Discussion

Following recent research results that showed that walking in virtual environments is effective for rehabilitation, this study aimed to analyze the correlations between HR, HRV, and SSQ scores with regard to motion sickness symptoms that can occur when walking, which is the most basic human movement, in a virtual environment. The HR difference was higher in the motion-sickness group than in the no-motion-sickness group, and the increase in HR difference was higher during HMD-assisted walking than during screen-assisted walking in the motion-sickness group. In addition, the RMSSD difference was lower in the motion-sickness group than in the no-motion-sickness group, and the decrease was greater during HMD-assisted walking than during screen-assisted walking in the motion-sickness group. The increase in the LF/HF ratio was greater during HMD-assisted walking than during screen-assisted walking in the motion-sickness group. The results of a regression analysis of the HR, HRV difference, and SSQ scores, which differentiated the motion-sickness and no-motion-sickness groups, showed that there was a high negative correlation between the SSQ scores and RMSSD difference.

Nalivaiko et al. (2015) asked 26 subjects to view a virtual roller coaster, evaluated the degree of motion sickness symptoms through an MSSQ survey, and measured their HRs and finger temperatures [17]. The results showed that there was little change in the HR for patients with nausea scores of less than five. However, the HR increased overall in patients with scores of five or greater (occurrence of cybersickness), and their finger temperatures were higher. Lin et al. (2018) presented three types of virtual environments to 25 subjects and asked the subjects to provide their motion sickness scores [7]. The changes in the symptoms were confirmed based on the HR and HRV of the group that experienced motion sickness and the group that did not experience motion sickness based on their motion sickness scores. As a result, it was found that the HR increased as the symptom scores increased, and the HRV was higher for the sickness group. Kim et al. (2022) presented VR content to 16 subjects, performed an SSQ survey, and measured the biosignals (blood pressure, cortisol, and HR) before and after viewing of the content [18]. It was found that the HR and cortisol levels were greatly increased in subjects who experienced cybersickness. Although the stimuli conditions differed from those in previous studies, the present study also found that the HR increased in the motion-sickness group when compared with that in the no-motion-sickness group during VR-assisted walking, which was the focus of this study, and the change in the HR of the motion-sickness group was greater during HMD-assisted walking than during screen-assisted walking. Particularly, it was found that the sympathetic nerves were activated more intensively during HMD-assisted walking than during screen-assisted walking in the motion-sickness group.

The RMSSD is the primary measure used to predict the beat-to-beat variability in the high-frequency band of the HRV; a larger RMSSD value indicates a more stable condition. Magaki and Vallance (2019) presented virtual environments in which 16 subjects attempted to locate five objects for five minutes under two conditions (using PCs and HMDs) and measured their HRs and SSQ scores [19]. Motion sickness symptoms in HMD-equipped walking were stronger, and there was a significant reduction in the mean value of NN intervals (NNMean), standard deviation of the NN intervals (SDNN), and RMSSD in the HMD conditions when compared with PC-equipped walking. Although these variables can be potential indicators of cybersickness, their correlation with the intensity of motion sickness remains uncertain. Gavgani et al. (2017) used the data from 12 subjects in their 20s and 30s who experienced a virtual roller coaster via HMDs (including scenes of embarking, moving forward, and moving backward) to determine the resulting degree of motion sickness symptoms (nausea) and measured the skin conductivity and ECGs [20]. Based on the results, they reported that the intensity of motion sickness was higher during forward-moving scenes than during backward-moving scenes, and the skin conductivity increased when motion sickness occurred, confirming the correlation between skin conductivity and motion sickness. In addition, there was an inverse correlation between the RMSSD and motion sickness symptoms, and the RMSSD was significantly reduced when the symptom level was high. In this study, RMSSD reflected the short-term variability in the HR and was a measurement of the parasympathetic nervous control of the heart. The RMSSD decreased to a greater extent in the motion-sickness group than in the no-motion-sickness group. Moreover, the difference was greater during HMD-assisted walking than during screen-assisted walking in the motion-sickness group. That is, it was found that the subjects became unstable during walking when motion sickness symptoms occurred, and the instability was greater when the stimuli were presented via an HMD than a screen.

The LF/HF ratio compares the relative power of the low-frequency band (0.04–0.15 Hz) to that of the high-frequency band (0.15–0.4 Hz). Under controlled conditions, this ratio can be used to estimate the ratio of sympathetic nervous system activity to parasympathetic nervous system activity [21]. Hsin et al. (2022) used HMDs to present 360° VR scenes to 28 subjects for ten minutes and assessed motion sickness using the HRV and SSQ questionnaire [22]. Ten patients had scores of 20 points or less, indicating mild simulator sickness, and the frequency of the SS symptoms increased as the screen speed increased. The SSQ scores showed a positive correlation with the very low frequency (VLF) band, and the VLF power was higher in the motion-sickness group than in the no-motion-sickness group, with a greater difference occurring in the 6–10 min interval. Kiryu and Iijima (2014) divided 15 male participants in their 20s into motion-sickness and no-motion-sickness groups according to the difference in the SSQ scores before and after repeated (five times) exposure to visual first-person-perspective motion sickness-inducing content [23]. The LF and HF time series data were acquired, and the LF values of 120% or above and HF values of below 80% were recorded as peak values. The motion sickness symptoms clearly increased as the frequency of the peaks increased. This study aimed to confirm that the LF/HF ratio is correlated with cybersickness by comparing the amount of change in the frequency of peaks and the amount of change in the LF/HF ratio at the peaks. The results showed that the LF/HF ratio increased when the frequency of the peak values increased, confirming the correlation between the LF/HF ratio and cybersickness. In addition, because the increase in the LF/HF ratio decreases over time when the frequency of the peaks decreases, it is expected that cybersickness is likely to have occurred because of a specific cybersickness-inducing trigger factor rather than simply because of prolonged exposure to virtual stimuli. Setiowati et al. (2019) conducted experiments with 15 participants wearing HMDs and performing driving simulation tasks for 15 min [24]. Subsequently, they conducted SSQ surveys in which the nausea score was negatively correlated with the LF/HF ratio; however, there was no statistically significant correlation with the RMSSD. In the present study, the LF/HF ratio increased to a greater extent in the motion-sickness group than in the non-motion-sickness group. Within the motion-sickness group, the difference was greater during HMD-assisted walking than during screen-assisted walking. Psychological stress due to motion sickness is significantly correlated with an increase in the LF/HF ratio [25]. LF power is related to sympathetic activity, and HF power is related to parasympathetic activity [26]. This indicates that the cybersickness induced in the participants increased the LF power and decreased the HF power. The increase in the LF power indicates that the sympathetic nervous system is dominant, which can be considered a bodily reaction to external stimuli.

This study aimed to find reliable data for determining the variable that can predict the presence of motion sickness among several variables measured on an ECG. The RMSSD had a high negative correlation with the SSQ scores; that is, the RMSSD scores decreased as the SSQ scores increased. It is possible that the RMSSD data can be used as a reliable variable that acts as an objective indicator to differentiate the motion-sickness group from the no-motion-sickness group during walking. RMSSD is one of the main variables used to analyze the short-term components of HRV. This variable represents the short-term heartbeat variability factors and is an indicator of the controlling capacity of the parasympathetic nerves. The reduction in RMSSD with an increase in the degree of motion sickness during walking in a virtual environment represents an inappropriate response by the autonomic nervous system function and appears to influence the cardiovascular system response.

Previous studies have measured motion sickness-related biosignals after constraining body movements. Hence, these studies cannot be compared with the present study, which measured motion sickness-related biosignals during walking. However, this study can be considered important fundamental research on rehabilitation incorporating virtual environments because it identified the HRV variables that can determine the presence of motion sickness when walking in VR. For example, if a patient’s RMSSD value changes abruptly, the intensity of rehabilitation activities can be automatically reduced to allow the patient to participate more comfortably in the rehabilitation process. Additionally, monitoring RMSSD can detect early signs of cybersickness in the VR environment. By promptly responding when symptoms of motion sickness begin to appear, i.e., by adjusting or pausing VR activities, it is possible to minimize patient discomfort and enhance the efficiency of rehabilitation. In the future, we plan to conduct experiments according to age and patient groups requiring rehabilitation to identify the variables that can determine the presence of motion sickness when walking in VR. In addition, we aim to classify the SSQ scores, which measure the degree of motion sickness, by item to identify the symptoms that appear most prominently during VR walking and apply this to the motion sickness evaluations. We also plan to examine the changes in walking pattern that occur during VR walking by group and age and analyze the correlation between the degree of motion sickness and changes in the walking pattern.

5. The Limitation of This Study

The limitation of this study is a small number of subjects, and further studies will be needed according to various age groups and genders. In addition, it was performed under two conditions, i.e., screen and HMD, and VR studies with various parameters applied will need to be further conducted. Certain indicators, such as the SSQ survey, ECG, HR, and HRV, were used for evaluating cyber motion sickness, but they may not have reflected all of the complex characteristics of cyber motion sickness. Further insight into other potential indicators (brain waves, skin conductivity, hormone levels, etc.) will be needed. Despite these limitations, this study contributes significantly to deepening the understanding of cyber motion sickness in a VR environment and suggesting the direction of future research. Based on the results of the study, studies targeting a wider population, experimental designs involving various VR environments and interactions, the exploration of additional physiological and psychological indicators, and studies evaluating long-term effects will be needed.