Artificial balance: restoration of the vestibulo-ocular reflex in humans with a prototype vestibular neuroprosthesis

Abstract

The vestibular system plays a crucial role in the multisensory control of balance. When vestibular function is lost, essential tasks such as postural control, gaze stabilization, and spatial orientation are limited and the quality of life of patients is significantly impaired. Currently, there is no effective treatment for bilateral vestibular deficits. Research efforts both in animals and humans during the last decade set a solid background to the concept of using electrical stimulation to restore vestibular function. Still, the potential clinical benefit of a vestibular neuroprosthesis has to be demonstrated to pave the way for a translation into clinical trials. An important parameter for the assessment of vestibular function is the vestibulo-ocular reflex (VOR), the primary mechanism responsible for maintaining the perception of a stable visual environment while moving. Here we show that the VOR can be artificially restored in humans using motion-controlled, amplitude modulated electrical stimulation of the ampullary branches of the vestibular nerve. Three patients received a vestibular neuroprosthesis prototype, consisting of a modified cochlear implant providing vestibular electrodes. Significantly higher VOR responses were observed when the prototype was turned ON. Furthermore, VOR responses increased significantly as the intensity of the stimulation increased, reaching on average 79% of those measured in healthy volunteers in the same experimental conditions. These results constitute a fundamental milestone and allow us to envision for the first time clinically useful rehabilitation of patients with bilateral vestibular loss.

Keywords: balance; rehabilitation; sensory neuroprostheses; vestibular implant; vestibulo-ocular reflex.

Figures

Figure 1

Vestibular implant prototype. An out-of-the-shelf cochlear implant (SONATA; MED-EL, Innsbruck, Austria) was modified in order to provide electrodes for stimulating vestibular structures. Three electrodes were taken out from the standard intracochlear array. Each of these “vestibular” electrodes was located on the distal tip of separate leads to allow implantation in the posterior, superior, and lateral ampullae.

Figure 2

Illustration of eye movement data processing. The figure presents eye movement data tracings for patient BVL1, gathered during the rotation experiments at a frequency of 1 Hz (modulation strength corresponding to 50% of the patient’s dynamic range). The panels on the left show data acquired during the system OFF experiments. The panels on the right show data gathered during the system ON experiments. Three steps are illustrated: raw eye position (e.g., before any processing was performed), processed eye position data (e.g., eye position data after blinks and quick eye movements >1000°/s2 were removed), and processed eye velocity data (e.g., obtained from the differentiation of the processed eye position data).

Figure 3

Vestibulo-ocular reflex responses of the three implanted patients to 30°/s peak-velocity sinusoidal rotations around the vertical axis in complete darkness at frequencies of 0.1, 0.25, 0.5, 1, and 2 Hz (columns). For each patient, the panels on the upper row show data gathered in the system OFF condition. The panels on the lower row show data gathered in the system ON condition. Solid lines represent the average cycle plots (±standard deviation, SD shown in dotted lines) of the horizontal angular velocity of the eye (red lines) and the head (blue lines). Note that at the lower frequencies eye movement recordings were polluted by random artifacts mainly due to the long duration of cycles at these frequencies (e.g., 10 s for 0.1 Hz rotations).

Figure 4

Vestibulo-ocular reflex response axes of the three implanted patients to 30°/s peak-velocity sinusoidal rotations around the vertical axis in complete darkness at frequencies of 0.1, 0.25, 0.5, 1, and 2 Hz (columns). Each individual panel shows average vertical versus horizontal eye position (red dots) gathered for each patient (rows) in the system ON condition (modulation strength corresponding to 50% of the dynamic range), at the different rotation frequencies tested (columns). The best linear fits to the data were calculated to estimate the axis of eye movements (angle with respect to the horizontal). Important eye dispersion was observed for the lower frequencies and the linear fits were not representative of the direction of eye movements. This was expected since the VOR response at these frequencies was almost absent and in these cases eye movement patterns were mainly random due to the long duration of cycles at these frequencies (e.g., 10 s for 0.1 Hz rotations). Otherwise eye movements were predominantly horizontal (maximum deviations of 2°, 21°, and 19° for BVL1, BVL2, and BVL3, respectively), and the axis remained almost the same despite the rotation frequency tested.

Figure 5

Mean phase-locked VOR gain (±SEM) versus rotation frequency of the three implanted patients. Two stimulation conditions are compared: system OFF (red plot) and system ON (blue plot; modulation strength corresponding to 50% of each patients dynamic range).

Figure 6

Mean phase-locked VOR gain (±SEM) versus stimulation condition of the three implanted patients. Three stimulation conditions are compared: system OFF and system ON with modulation strengths corresponding to 50 and 75% of each patient’s dynamic range. Rotations had a 30°/s peak-velocity sinusoidal profile (around the vertical axis) at 1 Hz.

Figure 7

Individual VOR performance recorded during the rotation experiments at 1 Hz. Each panel presents data gathered for one patient. Three conditions are compared: system OFF and system ON at two different modulation strengths (corresponding to 50 and 75% of each patient’s dynamic range, in parenthesis). Data were analyzed on a cycle-by-cycle basis. Box plots indicate median values, 25th and 75th percentile values (colored box) as well as 10th and 90th percentile values (error bars). The gray bar represents the mean ± SEM of the VOR gain measured in a group of five healthy subjects in the same experimental conditions.