Cybersickness Susceptibility: Preliminary fMRI Insights Informed by Psychological Metrics
Poster No:
2080
Submission Type:
Abstract Submission
Authors:
Laura Verdasco Menéndez1, Gang Li1, Yingying Huang1, Frank Pollick1
Institutions:
1University of Glasgow, Glasgow, Scotland
First Author:
Co-Author(s):
Introduction:
Beneficial applications of immersive VR like VR medicine and education are constrained by cybersickness. Assessing overlapping psychological and neural susceptibility factors could provide novel solutions. We investigated this by combining psychological and neural measures in two studies: questionnaire-based Study 1 (S1), and fMRI-based Study 2 (S2). S1 explored psychological susceptibility measure(s). From these, we hypothesized significant differences between susceptible and resistant participants in cybersickness insula resting state (RS) functional connectivity (FC). S2 was a pilot study of those hypotheses.
Methods:
S1:
71 participants (Age = 20-30) without VR/1st person videogame experience (>2 hours/month [1]) underwent a 30-min dynamic cybersickness induction. Induction used a 1st person cognitive task-integrated tunnel travel task [2], a HP Reverb G2 Omnicept Edition headset and a side-to-side rotating chair (< 0.2Hz [3]). Susceptible participants scored >11 in the 0-20 Fast Motion Sickness Scale (FMS [4]); otherwise, they were resistant. We assessed psychological factors interoception (Multidimensional Assessment of Interoceptive Awareness 2 (MAIA2) [5]), state anxiety (Current Anxiety Level Measure [6]), trait anxiety, stress, depression (Depression, Anxiety and Stress Scale [7]), and VR immersion (iGroup Presence Questionnaire [8]) with a between-subjects t-test per measure.
S2:
6 fMRI-experienced participants made up groups Susceptible (N = 3) and Resistant (N = 3). All completed 1) rest fMRI (10-min anatomical, 13-min eyes open RS), 2) cybersickness induction, and 3) cybersickness fMRI (13-min eyes open RS, 5-min anatomical). Induction was simplified into passive tunnel travel via Quest 2 headset without movement. Participants rated FMS for 20 minutes (Resistant) or until FMS>7 (Susceptible). Given the correlation between interoception and insula, we computed insular FC strength with Harvard-Oxford ROIs and produced General Linear Models with subjects random effects to assess group differences.
71 participants (Age = 20-30) without VR/1st person videogame experience (>2 hours/month [1]) underwent a 30-min dynamic cybersickness induction. Induction used a 1st person cognitive task-integrated tunnel travel task [2], a HP Reverb G2 Omnicept Edition headset and a side-to-side rotating chair (< 0.2Hz [3]). Susceptible participants scored >11 in the 0-20 Fast Motion Sickness Scale (FMS [4]); otherwise, they were resistant. We assessed psychological factors interoception (Multidimensional Assessment of Interoceptive Awareness 2 (MAIA2) [5]), state anxiety (Current Anxiety Level Measure [6]), trait anxiety, stress, depression (Depression, Anxiety and Stress Scale [7]), and VR immersion (iGroup Presence Questionnaire [8]) with a between-subjects t-test per measure.
S2:
6 fMRI-experienced participants made up groups Susceptible (N = 3) and Resistant (N = 3). All completed 1) rest fMRI (10-min anatomical, 13-min eyes open RS), 2) cybersickness induction, and 3) cybersickness fMRI (13-min eyes open RS, 5-min anatomical). Induction was simplified into passive tunnel travel via Quest 2 headset without movement. Participants rated FMS for 20 minutes (Resistant) or until FMS>7 (Susceptible). Given the correlation between interoception and insula, we computed insular FC strength with Harvard-Oxford ROIs and produced General Linear Models with subjects random effects to assess group differences.
Results:
S1:
There were significant results for 4 MAIA2 categories: Not Worrying (t(49.04)=2.38, p=.021), Emotional Awareness (t(39.79=-2.36, p=.023), and Noticing (t(54.33)=-2.23, p=.030). This indicates that, among factors like presence or trait anxiety, stress, or depression, only interoception significantly affects susceptibility. Based on this, we designed fMRI-based S2 [9].
S2:
There were two significant right insula FC clusters (T(4)>8.61, k>15) (Figure 1). Cluster 1 was in the inferior temporal gyrus and temporal fusiform cortex (Centre(MNI): +44, -34, -16, size(3x3x3mm)=19, size pFWE=.081, size pFDR<.05, peak pFWE=1). Cluster 2 was in the right middle frontal gyrus and right precentral gyrus (Centre: +34, +10, +42, size=15, size pFWE=.231, size pFDR=.049, peak pFWE=1). These suggest susceptible participants' right insula FC significantly decreased during cybersickness.
There were three significant left insula FC clusters (T(4)>8.61, k>14) (Figure 2). Cluster 1 was in the right pre- and postcentral gyrus (Centre(MNI): +40, -20, +58, size(3x3x3mm)=34, size pFWE<.001, size pFDR<.001, peak pFWE=1). Cluster 2 was in the right frontal pole (Centre: +40, +52, +24, size=16, size pFWE= .107, size pFDR=.016, peak pFWE=1). Cluster 3 was in the precuneus cortex (Centre: -02, -56, +62, size=14, size pFWE=.196, size pFDR=.022, peak pFWE=1). These suggest susceptible participants' left insula FC significantly decreased during cybersickness.
There were significant results for 4 MAIA2 categories: Not Worrying (t(49.04)=2.38, p=.021), Emotional Awareness (t(39.79=-2.36, p=.023), and Noticing (t(54.33)=-2.23, p=.030). This indicates that, among factors like presence or trait anxiety, stress, or depression, only interoception significantly affects susceptibility. Based on this, we designed fMRI-based S2 [9].
S2:
There were two significant right insula FC clusters (T(4)>8.61, k>15) (Figure 1). Cluster 1 was in the inferior temporal gyrus and temporal fusiform cortex (Centre(MNI): +44, -34, -16, size(3x3x3mm)=19, size pFWE=.081, size pFDR<.05, peak pFWE=1). Cluster 2 was in the right middle frontal gyrus and right precentral gyrus (Centre: +34, +10, +42, size=15, size pFWE=.231, size pFDR=.049, peak pFWE=1). These suggest susceptible participants' right insula FC significantly decreased during cybersickness.
There were three significant left insula FC clusters (T(4)>8.61, k>14) (Figure 2). Cluster 1 was in the right pre- and postcentral gyrus (Centre(MNI): +40, -20, +58, size(3x3x3mm)=34, size pFWE<.001, size pFDR<.001, peak pFWE=1). Cluster 2 was in the right frontal pole (Centre: +40, +52, +24, size=16, size pFWE= .107, size pFDR=.016, peak pFWE=1). Cluster 3 was in the precuneus cortex (Centre: -02, -56, +62, size=14, size pFWE=.196, size pFDR=.022, peak pFWE=1). These suggest susceptible participants' left insula FC significantly decreased during cybersickness.
Conclusions:
Informed by S1 findings, we designed S2. S1 showed interoception is a susceptibility factor, thus S2 hypothesized insula FC during cybersickness would differ significantly between groups. As predicted, susceptible participants' right insula showed decreased FC with visual processing, internal stimuli attention, and motion areas. Additionally, susceptible participants' left insula showed decreased FC with cognition, mental imagery, and symptom awareness areas. Larger studies could assess replicability.
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural)
fMRI Connectivity and Network Modeling
Motor Behavior:
Visuo-Motor Functions 1
Motor Behavior Other
Perception, Attention and Motor Behavior:
Perception: Multisensory and Crossmodal 2
Keywords:
FUNCTIONAL MRI
Perception
Other - Cybersickness
1|2Indicates the priority used for review
Provide references using author date format
[2] Pöhlmann K. M. T., Li G., McGill M., Pollick F., and Brewster S. (2023), ‘Can Gender and Motion Sickness Susceptibility Predict Cybersickness in VR?’, 2023 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW), Shanghai, China, 2023, pp. 277-282.
[3] Golding JF, Mueller AG, Gresty MA (2001). ’A motion sickness maximum around the 0.2 Hz frequency range of horizontal translational oscillation’, Aviat Space Environ Med, vol. 72, no. 3, pp. 88-92.
[4] Keshavarz, B., & Hecht, H. (2011). ‘Validating an efficient method to quantify motion sickness’, Human Factors, vol. 53, no. 4, pp. 415 – 426.
[5] Mehling WE, Acree M, Stewart A, Silas J, Jones A (2018), ‘The Multidimensional Assessment of Interoceptive Awareness, Version 2 (MAIA-2)’, PLoS ONE, vol. 13, no. 12.
[6] Marris, Melissa & Sladyk, Karen & Pierre, Beverly & Dey, Elizabeth. (2017), ‘Reliability and Validity of the Current Anxiety Level Measure’, Occupational Therapy in Mental Health, vol. 33, pp. 1-8.
[7] Lovibond, S. H., & Lovibond, P. F. (1995). ‘Depression Anxiety Stress Scales (DASS--21, DASS--42)’ [Database record], APA PsycTests.
[8] Schubert, T. (2003). ‘The sense of presence in virtual environments: A three-component scale measuring spatial presence, involvement, and realness’, Zeitschrift für Medienpsychologie, vol- 15, pp. 69-71.
[9] Craig AD (2002). ‘How do you feel? Interoception: the sense of the physiological condition of the body’. Nat Rev Neurosci, vol. 3, no. 8, pp. 655-66.