Biologically inspired computational models relating vection, optokinetic nystagmus (OKN) and visually induced motion sickness (VIMS)

Motivated by recent debates on the role of vection and optokinetic nystagmus (OKN) in VIMS generation, two biologically inspired computational models are developed to simulate the connections between vection and VIMS (model one), and the connections between OKN and VIMS (model two), respectively. Both models are new and consistent with their respective VIMS theories and neuroscience knowledge concerning VIMS. Results of simulation of models one and two successfully reproduce sickness severity levels that are comparable to the empirical data published in Stern et al. (1990) and Webb and Griffin (2002), respectively. A literature review reveals a lack of empirical study reporting the isolated effects of vection and OKN on ratio scale data of VIMS severity. Consequently, two empirical experiments are conducted to validate the two models, respectively. Experiment one examines the effects of circular vection (CV) on VIMS severity in the absence of OKN while watching visual stripe patterns rotating in the yaw axis at different velocities. Experiment two examines the effect of optokinetic nystagmus (OKN) on VIMS severity in the absence of vection while watching visual stripe patterns rotating in the yaw axis at different velocities. Both experiments one and two report a significant main effect of rotation velocity of the optokinetic stimuli on VIMS severity. Experiment one reports a linear relationship between the vection velocity and VIMS severity. Experiment two reports a nonlinear relationship between the foveal retinal slip velocity and VIMS severity. Such results indicate that either vection alone or OKN alone can induce VIMS and their effects change significantly as the visual stimuli velocity changes. Results of model validation indicate that both models can simulate the average VIMS severity levels as functions of rotation velocities of the stripe patterns. In addition, sensitivity analyses are conducted to identify the model parameters contributing to inter-subject variations in VIMS severity. This enables both models to simulate not only the mean VIMS severity levels, but also the inter-subject variations. Model one and model two can simulate inter-subject variations in VIMS severity by manipulating inter-subject variations in three vection related measures (latency of vection onset, vection build-up time, the steady state of vection velocity) and three OKN related measures (OKN slow-phase velocity, OKN gain, and foveal retinal slip velocity), respectively. Such results are consistent with the understanding that both vection and OKN play a significant role in VIMS generation. In addition, both experiments report significant correlation among the three VIMS severity measures: (i) 7-point nausea rating, (ii) SSQ nausea score, and (iii) nausea ratio scale data collected by free-modulus magnitude estimation. It is proposed the free-modulus magnitude estimation, as a supplement to the two ordinal scales for measuring VIMS severity, can benefit future VIMS studies