Augmentation of Sensorimotor Adaptability Using Stochastic Resonance Technologies

We have found that the application of low level, noisy electrical current to the vestibular system via electrodes placed behind each ear can improve human vestibular perception of low level physical motion stimuli. We believe that this improvement in sensory performance is due to the exhibition of stochastic resonance (SR), in which added noise enables better transfer of information through the nonlinear system. First we investigated the effect of stochastic vestibular stimulation (SVS) on direction recognition (DR) thresholds (i.e. the smallest motion that a person can reliably perceive leftward from rightward motion while in darkness). We used 0.2 Hz head-centered upright roll tilt away from the upright to elicit responses from both vestibular organs - the semicircular canals and the otoliths. We found that SR was exhibited subjectively in 9/12 subjects and statistically in 6/12 subjects, with percent changes in DR threshold as large as -47%. Further, the relationship between DR thresholds and SVS level was consistent with the characteristic SR pseudo bell shaped curve within individual subjects. To provide further support, this relationship was modeled using theoretical SR equations originating from an entirely different domain: the study of Earth’s changes in glacial states throughout it’s history as they relate to the changes in Earth’s orbital eccentricity. Past studies of SR have reported that SR was not exhibited in all subjects tested. In our subject groups there was also less than 100% SR exhibition, however our proportion of subjects that exhibited SR in upright roll tilt was at least as high and often larger than previous SR studies that tested human subject performance. In order to better understand the likelihood of our data indicating the exhibition of SR when in fact there was no underlying SR and conversely of underlying SR existing but not being measured by our methods, we used numerical simulations of our experiment protocol to create a large dataset of simulated subjects with varying (controlled) levels of underlying SR. A comparison between our experimental data and the simulation data further solidified the exhibition of SR in the upright roll tilt motion with SVS application, even when not every subject exhibited SR. Whether galvanic vestibular stimulation (GVS) differentially stimulates the semicircular canals or otoliths has been a topic of debate which to date is not fully settled. Whether SVS differentially affects the two vestibular organs is also unclear and relatively unstudied. Therefore, in a second experiment, we investigated the effect of SVS on supine roll rotation and inter-aural translation DR thresholds, which primarily stimulate the semicircular canals and otoliths, respectively. By again comparing our experimental results to our simulated dataset, we found essentially no SR exhibition in the supine roll rotation data set (subjectively in 3/12 subjects and statistically in 1/12 subjects). We found less ambiguous SR exhibition in the inter-aural translation measurements (subjectively in 7/11 subjects and statistically in 6/11 subjects), with proportions of SR exhibitors comparable to the upright roll tilt subject group from Experiment 1. These results suggest that the otoliths, the vestibular organs that respond to gravity and linear accelerations, are an integral component to vestibular perceptual SR exhibition. The results of our second experiment suggest that although SVS is simply a lower, noisier version of GVS, the effects of SVS and GVS on vestibular perception may not be directly related. Finally, we aimed to study if SVS could benefit human sensorimotor performance in a task that is operationally relevant in the aerospace field. We studied the effect of a zero, low, and high level of SVS on the ability to null out whole body motion roll disturbances using a manual control bar, a task relevant to piloting. We found that low level SVS (same ± 300 µA level that caused minimal DR thresholds in many subjects in Experiment 1) helped subjects to maintain a smaller range of motion when the input disturbance was near sensory threshold (statistically significant for the group mean, corresponding to a decrease in position variability of 21% and for 2/8 individual subjects with a maximum decrease in position variability of 41%), but not over a longer period of time in which the disturbance was larger and more variable. The high level of SVS tested had no measurable effect on mean manual control performance. Since it is well known that vestibular system function is highly dependent on the frequency of the motion stimuli, we also did a frequency analysis of the manual control data. With the low level of SVS applied, 6/8 subjects were able to null out motions at a mid-frequency (0.2-0.26Hz) that with 0 SVS, was out of the controllable frequency range. The difference in the Scalar Performance Metric (SPM), a measure of nulling performance at a specific frequency or set of frequencies, was statistically significant for the group mean corresponding to a 7 times increase in SPM with the application of low level SVS than without it. These results suggest that low-level SVS can extend the frequency range in which upright roll tilt motions can be sensed and nulled out in a manual control task. A few individual subjects also showed unique frequency-specific performance changes between SVS levels that were inconsistent with the group means, suggesting that subject-specific responses to SVS may be also depend on the frequency of the physical motion stimulus. The results of this thesis are consistent with the concept that SVS is able to extend the operating range of the vestibular perceptual system in some individuals. As such, the future of SVS applications may be broad and far reaching. Future studies should focus on better understanding inter-individual differences in the response to SVS and aim to increase the proportion of subjects that exhibit SR and the reliability of SR repeatability. These practical aspects of SVS are important in ensuring that SVS continue to be considered for future implementation in the aerospace or clinical fields.