(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



A prosthesis utilizing natural vestibular encoding strategies improves sensorimotor performance in monkeys [1]

['Kantapon Pum Wiboonsaksakul', 'Department Of Biomedical Engineering', 'Johns Hopkins University School Of Medicine', 'Baltimore', 'Maryland', 'United States Of America', 'Kavli Neuroscience Discovery Institute', 'Johns Hopkins University', 'Dale C. Roberts', 'Charles C. Della Santina']

Date: 2022-09

Sensory pathways provide complex and multifaceted information to the brain. Recent advances have created new opportunities for applying our understanding of the brain to sensory prothesis development. Yet complex sensor physiology, limited numbers of electrodes, and nonspecific stimulation have proven to be a challenge for many sensory systems. In contrast, the vestibular system is uniquely suited for prosthesis development. Its peripheral anatomy allows site-specific stimulation of 3 separate sensory organs that encode distinct directions of head motion. Accordingly, here, we investigated whether implementing natural encoding strategies improves vestibular prosthesis performance. The eye movements produced by the vestibulo-ocular reflex (VOR), which plays an essential role in maintaining visual stability, were measured to quantify performance. Overall, implementing the natural tuning dynamics of vestibular afferents produced more temporally accurate VOR eye movements. Exploration of the parameter space further revealed that more dynamic tunings were not beneficial due to saturation and unnatural phase advances. Trends were comparable for stimulation encoding virtual versus physical head rotations, with gains enhanced in the latter case. Finally, using computational methods, we found that the same simple model explained the eye movements evoked by sinusoidal and transient stimulation and that a stimulation efficacy substantially less than 100% could account for our results. Taken together, our results establish that prosthesis encodings that incorporate naturalistic afferent dynamics and account for activation efficacy are well suited for restoration of gaze stability. More generally, these results emphasize the benefits of leveraging the brain’s endogenous coding strategies in prosthesis development to improve functional outcomes.

Competing interests: I have read the journal’s policy and the authors of this manuscript have the following competing interests: C.C.D.S. is an inventor of pending and awarded patents related to technologies discussed in this manuscript, and he holds an equity interest in and is the CEO/CSO of Labyrinth Devices, LLC. The terms of this arrangement are managed in accordance with Johns Hopkins University policies on conflict of interest.

To address this question, here, we directly tested whether implementing the natural response dynamics of vestibular afferents improves prosthesis performance in a nonhuman primate model. We integrated these response dynamics into the mapping between head motion and pulse rate delivered by a vestibular prosthesis targeting the ampullary nerve innervating the horizontal semicircular canal [ 16 – 19 ]. Quantification of the resulting eye movements produced by the VOR provided a direct measure of prosthesis performance. Overall, we found that biomimetic mappings that accounted for the brain’s endogenous dynamic coding strategies produced more temporally accurate VOR eye movements than those that did not. Further exploration of the parameter space (i.e., the variables that control the mapping of head movement into stimulation rate) revealed that incorporating even more extreme tunings than those naturally displayed by vestibular afferents provided unnatural phase advances, as well as undesirable nonlinear gain saturation. Trends were comparable for stimulation encoding virtual versus physical head rotations, with gains enhanced in the latter case. Using computational methods, we then demonstrated that the same model could explain the eye movements evoked by both sinusoidal and transient stimulation and that a stimulation efficacy substantially less than 100% could account for our results—i.e., each stimulation pulse does not always evoke an afferent action potential. Thus, taken together, our results show that mappings incorporating naturalistic afferent dynamics and stimulation efficacy improve the restoration of gaze stability. More broadly, these results also underscore the benefits of leveraging the brain’s natural coding strategies to improve functional outcome of sensory prostheses.

The vestibular system detects our head motion relative to space. Rotational head motion is sensed by the 3 semicircular canals and linear acceleration is sensed by the 2 otolith organs. In turn, this information is used to generate essential stabilizing reflexes and complex motor synergies that control gaze and posture, in addition to providing us with our subjective sense of motion and orientation. There are many reasons why the vestibular system is uniquely well suited for the development of neural prostheses for the restoration of sensory function. First, the 3 axes of head rotation are encoded by 3 separate semicircular canals, each innervated by a distinct nerve bundle. This sensor structure allows for targeted stimulation of each of the associated orthogonal rotational axes. Second, vestibular afferent response dynamics have been extensively studied and are well understood ([ 10 – 13 ]; also reviewed in [ 14 ]). Third, the central pathways mediating essential vestibular reflexes, such as the vestibulo-ocular reflex (VOR), which serves to stabilize gaze during head motion, are simple—comprising 3-neuron circuits in their most direct forms (reviewed in [ 14 ]). As a result, the vestibulo-motor transformations generating these rapid stereotyped behavioral responses can be objectively quantified and utilized as a direct measure of prosthesis performance [ 15 – 17 ]. Despite these advantages, whether leveraging the brain’s endogenous coding strategies can improve vestibular prosthesis performance surprisingly remains unknown.

Even the simplest sensorimotor transformations require precise neural dynamics. For example, the visuomotor transformations required to produce voluntary saccadic and smooth pursuit eye movements must precisely account for the biomechanical properties of the extraocular muscles and orbital connective tissue (reviewed in [ 1 , 2 ]). Recent developments of sensory prostheses have focused on mimicking the response dynamics of peripheral sensors/afferent systems to improve the functional outcomes of retinal [ 3 ] and cochlear [ 4 , 5 ] implants, as well as prostheses aimed at restoring tactile sensation [ 6 – 9 ]. However, complex sensor physiology, limited number of electrodes, and nonspecific stimulation have proven to be a challenge to the development of biomimetic prostheses in these sensory systems.

Results

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001798

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/