Linear and non-linear measures of postural control predict individual variations in illusions of self-motion

Abstract

Background

Strong illusions of self-motion (vection, sometimes known as the “train illusion”) can be generated by large-field visual displays. These illusions can be highly compelling for some people, but there are large variations between individuals in vection strength. Do these variations depend, at least in part, on the extent to which people rely on vision to control their postural stability? We used measures of postural stability (quiet stance with eyes open and closed, and visually-induced sway) to examine the relationship between visual control of posture and self-motion illusions.

Methods

Using a Bertec balance plate in a brightly-lit room, we measured excursions of the centre of foot presure (CoP) over a 60-second period with eyes open and with eyes closed, for 13 participants. We also collected measures of postural sway (changes in anterior-posterior CoP) in response to large expanding and contracting optic flow stimuli while standing on the plate. Subsequently, we collected vection strength ratings for the same optic flow displays while participants were seated, using both verbal ratings and online throttle measures.

Results

We examined both linear (eyes open/ eyes closed sway area ratio and visually-induced sway magnitude) and non-linear (Recurrence Quantification Analysis) measures and their relationship to vection strength. Each of these measures significantly predicted individual variations in vection strength, and the measures were not correlated with each other. Interestingly, the non-linear measures showed much stronger predictions for both expanding and contracting vection ratings.

Discussion

Visual control of posture is a useful index for predicting individual susceptibility to illusions of motion. This could have implications for computer gaming, flight simulations, and future studies of vection. In addition, I will discuss the use of non-linear measures of scale-free dynamics in time series data, which can be extended to other applications such as EEG and heart rate variability.