Disentangling the visual, motor and representational effects of vestibular input

Abstract

The body midline provides a basic reference for egocentric representation of external space. Clinical observations have suggested that vestibular information underpins egocentric representations. Here we aimed to clarify whether and how vestibular inputs contribute to egocentric representation in healthy volunteers. In a psychophysical task, participants were asked to judge whether visual stimuli were located to the left or to the right of their body midline. Artificial vestibular stimulation was applied to stimulate the vestibular organs. We found that artificial stimulation of the vestibular system biased body midline perception. Importantly, no effect was found on motor effector selection. We also ruled out additional explanations based on allocentric visual representations and on potential indirect effects caused by vestibular-driven movements of the eyes, head and body. Taken together our data suggest that vestibular information contributes to computation of egocentric representations by affecting the internal representation of the body midline.

Introduction

Judging the position of external objects relative to the body is essential for interacting with the external environment. Egocentric representations describe the external world as experienced from an individual's location, according to the current spatial configuration of their body (Jeannerod & Biguer, 1987). Consider, for example, a tennis player who must quickly select a forehand or backhand shot based on the ball location relative to their body. A coherent and rapid response to the approaching ball requires combining perceptual information about the ball's trajectory relative to the player with information about the player's ever-changing posture and gaze. Such egocentric representations are thought to be essential in representing the world in relation to oneself (Bermúdez, 2005, Bermúdez, 2011, Cassam, 2011, Pafel et al., 1998).

The body midline may provide a basic reference for egocentric representation of external space (Jeannerod & Biguer, 1987). Everyday descriptions of spatial locations frequently begin with “on the left …” or “on the right …”. The subjective body midline is considered the internal representation of the plane that divides the body in two equal left and right parts (Bowers and Heilman, 1980, Jeannerod and Biguer, 1987). It remains unclear whether the subjective body midline co-ordinates are a static stored representation reflecting primarily semantic knowledge about body morphology, or rather a dynamic, continuously updated sensory datum, perhaps reflecting balance between afferent signals from lateralized receptor organs (left and right eyes, ears etc.), across changing body posture and orientation (Critchley, 1953).

Visual, auditory, somatosensory, proprioceptive and vestibular inputs could all contribute to representing the body midline (Blouin et al., 1996, Blouin et al., 1998, Jeannerod, 1988). Vestibular signals seem to be particularly relevant (Bonnier, 1905, Lhermitte, 1952, Schilder, 1935, Vallar and Papagno, 2003, Vallar and Rode, 2009). The vestibular system comprises the semicircular canals that encode rotational movements, and the otolith organs that encode translational accelerations, including the current orientation of the head relative to the gravitational vertical. Both semicircular canals and otolith organs constantly provide afferent information regarding body orientation and body movement. Since the vestibular organs on each side of the body act in a push/pull manner, a balance between vestibular signals can guide representation of the body midline. For example, a linear acceleration that produces identical signals from both otolith organs must correspond to movement aligned with the body midline, either in the straight-ahead or up-down direction. Similarly, any head rotation away from alignment with the body midline should cause equal and opposite changes in firing rate from the horizontal canals on both sides of the body. Thus, vestibular information is crucial to determine the location of environmental objects in respect to the body (Clément et al., 2009, Clément et al., 2012, Villard et al., 2005), and to specify the body midline itself.

Several clinical observations have suggested that vestibular information underpins egocentric representations. Patients with unilateral spatial neglect showed a deviation to the ipsilateral half of space when they were requested to point to an imaginary location in space straight ahead from their body midline (Heilman, Bowers, & Watson, 1983). Critically, artificial stimulation of the vestibular system influenced this pointing error: left cold caloric vestibular stimulation temporarily reduced the rightward pointing bias characteristic of patients with left-side neglect. This suggests that vestibular inputs contribute to the subject's mental representation of space and subjective body orientation (Karnath, 1994). However, most of these studies used motor pointing responses to estimate perceptual estimates of the body midline. That is, they assumed that the impairment arose at the level of representation of the body midline, but they could not formally exclude the possibility that vestibular stimulation affected the motor pointing response, or some purely visual element of the experiment. Here we aimed to clarify whether and how vestibular inputs contribute to egocentric spatial representation in healthy volunteers. We have systematically investigated which processing stages along the visual-motor processing chain are modulated by vestibular signals. This method allowed us to dissociate vestibular effects on visual perception and on motor action from effects on spatial representation, seemingly for the first time.

Binaural bipolar Galvanic Vestibular Stimulation (GVS) was used to non-invasively stimulate the vestibular receptors (Fitzpatrick & Day, 2004). An anode and cathode are placed on the left and right mastoid, or vice versa. Perilymphatic cathodal currents depolarize the trigger site and lead to excitation, whereas anodal currents hyperpolarize it resulting in inhibition (Goldberg, Smith, & Fernandez, 1984). GVS causes polarity-dependent modulation of sensory and cognitive functions (Utz, Dimova, Oppenländer, & Kerkhoff, 2010). Importantly, these behavioural effects are consistent with neuroimaging evidence revealing asymmetrical cortical vestibular projections in the non-dominant hemisphere (Dieterich et al., 2003).

We hypothesized that vestibular information might play a role in shaping the online perception of the body midline, and thus contribute a basic reference for egocentric spatial representation. Accordingly, we dissociated the vestibular contributions to egocentric spatial representations from those to motor responses (Experiment 1). In a second experiment, we investigated whether the GVS-induced bias on body midline could be explained by biases in visual perception, particularly in visual allocentric representation, and found that it could not (Experiment 2). Finally, we showed that the effects of GVS on egocentric representation were qualitatively distinct from the effects of GVS on gaze location (Experiments 3 and 4).