Title: “Physiological Computations Underlying Our Internal Representation of Vestibular Self-motion”
Thesis Supervisors: Dr. Jean-Sébastien Blouin & Dr. Robert Boushel
Committee Member: Dr. Christopher Dakin
Chair: Dr. Cheryl L. Wellington
External Examiner: Dr. Raymond Reynolds
University Examiners: Dr. Miriam Spering & James Enns
Abstract:
Humans rely on sensing for survival. The principles implemented by the brain to interpret sensory information from encoded neural impulses remain open. The vestibular sense provides information about self-motion and orientation, which in part allows us to properly perceive motion in space and to navigate the surrounding environment. The purpose of this thesis was to advance the current knowledge of how the brain processes sensory cues of motion. Given that the central integration properties of vestibular information are difficult to reveal through everyday experiences, we used experimentally controlled sensory stimuli to probe neurophysiological and psychophysical behaviours. Studies presented here emerge from a first principles approach looking at integration of signals from primary vestibular afferents using a novel sampling-based dynamical inference model. Experiment 1 characterized perceptual self-motion during and following constant vestibular stimulation and contextualized how percepts arise from the integration of differently adapting primary vestibular afferents based on the brain expectation about one’s self-motion. Experiment 2 examined how different vestibular end-organs encoding the same whole-body motion leads to dissimilar self-motion percepts following central integration. We showed that vestibular otoliths provide the brain additional information of self-rotation, a role primarily attributed to the vestibular semicircular canals. Experiments 3 and 4 together evaluated electrical vestibular stimulation for probing the vestibular system in humans. Here, we used current physiological data of primary vestibular afferent responses to both mechanical and electrical stimuli to predict and evaluate their equivalence in central processing in humans. Subsequently, a real-time solution for human-in-the-loop vestibular sense modulation using electrical vestibular stimulation during active head movements was developed using our identified equivalency. Altogether, the findings of this thesis highlight that the brain may exert a top-down influence in the form of prior expectations during sensory integration. Likewise, we show that equivalent perceptual and oculomotor behaviours can be elicited by matching the predicted response of primary vestibular afferents. The novel work presented here offers new theoretical and applied principles about vestibular processing, and the sensory integration process at large. More importantly, this may lead to novel and testable hypotheses for future studies to build upon in both fundamental and clinical areas.