Sensory substitution in bilateral vestibular a-reflexic patients

Abstract

Patients with bilateral vestibular loss have balance problems in darkness, but maintain spatial orientation rather effectively in the light. It has been suggested that these patients compensate for vestibular cues by relying on extravestibular signals, including visual and somatosensory cues, and integrating them with internal beliefs. How this integration comes about is unknown, but recent literature suggests the healthy brain remaps the various signals into a task-dependent reference frame, thereby weighting them according to their reliability. In this paper, we examined this account in six patients with bilateral vestibular a-reflexia, and compared them to six age-matched healthy controls. Subjects had to report the orientation of their body relative to a reference orientation or the orientation of a flashed luminous line relative to the gravitational vertical, by means of a two-alternative-forced-choice response. We tested both groups psychometrically in upright position (0°) and 90° sideways roll tilt. Perception of body tilt was unbiased in both patients and controls. Response variability, which was larger for 90° tilt, did not differ between groups, indicating that body somatosensory cues have tilt-dependent uncertainty. Perception of the visual vertical was unbiased when upright, but showed systematic undercompensation at 90° tilt. Variability, which was larger for 90° tilt than upright, did not differ between patients and controls. Our results suggest that extravestibular signals substitute for vestibular input in patients' perception of spatial orientation. This is in line with the current status of rehabilitation programs in acute vestibular patients, targeting at recognizing body somatosensory signals as a reliable replacement for vestibular loss.

Keywords: Bilateral vestibular a‐reflexia; internal models; multisensory integration; spatial orientation; verticality perception.

Figure 1

(A) Schematic representation of the multisensory integration model by Clemens et al. (2011) In the SBT task, body somatosensory signals provide direct information about orientation of body orientation in space, whereas the otoliths provide indirect information about the body orientation in space by taking into account the head-on-body information provided by neck proprioception. Similarly, in the SVV task otoliths provide direct information while body somatosensory signals combined with neck proprioceptors provide indirect information about head orientation in space. Both direct and indirect pathways are optimally combined for best performance on the tasks. Note that in the SVV task prior information about our head-in-space influences these pathways. Secondly, to compute the SVV, the brain also uses estimates of eye-in-head orientation (ocular counterroll) and line orientation on the retina (not shown here). The red arrows indicate information pathways that are lost in bilateral vestibular patients. (B) SBT task: subject has to indicate whether body orientation is clockwise (CW) or counterclockwise (CCW) of a certain reference orientation (dashed line) (C) SVV task: subjects are rotated to either upright or 90° RED and have to indicate whether a luminous line flashed in front of them is oriented clockwise (CW) or counterclockwise (CCW) of gravity.

Figure 2

SBT and SVV performance for a typical control and patient. SBT: proportion of clockwise responses P(CW) is plotted against body orientation centered on the two reference angles (0° or 90° RED). SVV: proportion of clockwise responses P(CW) is plotted against line orientation with respect to the vertical when the body is either upright or at 90° RED.

Figure 3

Mean bias and response variability at the upright and 90° roll tilt in both the SBT and SVV task. Error bars denote the standard error across subjects. * indicates P < 0.05, **P < 0.01, and ***P < 0.001.