Slow changing postural cues cancel visual field dependence on self-tilt detection.

Interindividual differences influence the multisensory integration process involved in spatial perception. Here, we assessed the effect of visual field dependence on self-tilt detection relative to upright, as a function of static vs. slow changing visual or postural cues. To that aim, we manipulated slow rotations (i.e., 0.05° s(-1)) of the body and/or the visual scene in pitch. Participants had to indicate whether they felt being tilted forward at successive angles. Results show that thresholds for self-tilt detection substantially differed between visual field dependent/independent subjects, when only the visual scene was rotated. This difference was no longer present when the body was actually rotated, whatever the visual scene condition (i.e., absent, static or rotated relative to the observer). These results suggest that the cancellation of visual field dependence by dynamic postural cues may rely on a multisensory reweighting process, where slow changing vestibular/somatosensory inputs may prevail over visual inputs.

How do visual and postural cues combine for self-tilt perception during slow pitch rotations?

Self-orientation perception relies on the integration of multiple sensory inputs which convey spatially-related visual and postural cues. In the present study, an experimental set-up was used to tilt the body and/or the visual scene to investigate how these postural and visual cues are integrated for self-tilt perception (the subjective sensation of being tilted). Participants were required to repeatedly rate a confidence level for self-tilt perception during slow (0.05°·s(-1)) body and/or visual scene pitch tilts up to 19° relative to vertical. Concurrently, subjects also had to perform arm reaching movements toward a body-fixed target at certain specific angles of tilt. While performance of a concurrent motor task did not influence the main perceptual task, self-tilt detection did vary according to the visuo-postural stimuli. Slow forward or backward tilts of the visual scene alone did not induce a marked sensation of self-tilt contrary to actual body tilt. However, combined body and visual scene tilt influenced self-tilt perception more strongly, although this effect was dependent on the direction of visual scene tilt: only a forward visual scene tilt combined with a forward body tilt facilitated self-tilt detection. In such a case, visual scene tilt did not seem to induce vection but rather may have produced a deviation of the perceived orientation of the longitudinal body axis in the forward direction, which may have lowered the self-tilt detection threshold during actual forward body tilt.

Influence of postural constraints on eye and head latency during voluntary rotations.

Redirecting gaze towards new targets often requires not only eye movements, but also synergistic rotations of the head, trunk and feet. This study investigates the influence of postural constraints on eye and head latency during voluntary refixations in the horizontal plane in 14 normal subjects. Three postural conditions were presented, (1) sitting in a chair using only eye and head movements, (2) standing without feet movements and (3) standing with feet movement. Head-eye reorientations towards eccentric un-predictable locations were performed towards ±45° and ±90° targets and back towards a central, spatially predictable target. Results showed that postural constraints affected eye latency but only when subjects knew the future location of the target (recentering "return" trials). Specifically, relatively longer eye latencies were observed when subjects had to turn their feet back towards the predictable central target. These findings suggest that the additional CNS processing required to reduce degrees of freedom during predictive motion introduces delays to the eye movement in order to efficiently assemble the components of a new motor synergy.

Spatial localization investigated by continuous pointing during visual and gravitoinertial changes.

In order to accurately localize an object, human observers must integrate multiple sensory cues related to the environment and/or to the body. Such multisensory integration must be repeated over time, so that spatial localization is constantly updated according to environmental changes. In the present experimental study, we examined the multisensory integration processes underlying spatial updating by investigating how gradual modifications of gravitoinertial cues (i.e., somatosensory and vestibular cues) and visual cues affect target localization skills. These were assessed by using a continuous pointing task toward a body-fixed visual target. The "single" rotation of the gravitoinertial vector (produced by off-axis centrifugation) resulted in downward pointing errors, which likely were related to a combination of oculogravic and somatogravic illusions. The "single" downward pitch rotation of the visual background produced an elevation of the arm relative to the visual target, suggesting that the rotation of the visual background caused an illusory target elevation (induced-motion phenomenon). Strikingly, the errors observed during the "combined" rotation of the visual background and of the gravitoinertial vector appeared as a linear combination of the errors independently observed during "single" rotations. In other words, the centrifugation effect on target localization was reduced by the visual background rotation. The observed linear combination indicates that the weights of visual and gravitoinertial cues were similar and remained constant throughout the stimulation.