Characterization of Cochlear, Vestibular and Cochlear-Vestibular Electrically Evoked Compound Action Potentials in Patients with a Vestibulo-Cochlear Implant

Abstract

The peripheral vestibular system is critical for the execution of activities of daily life as it provides movement and orientation information to motor and sensory systems. Patients with bilateral vestibular hypofunction experience a significant decrease in quality of life and have currently no viable treatment option. Vestibular implants could eventually restore vestibular function. Most vestibular implant prototypes to date are modified cochlear implants to fast-track development. These use various objective measurements, such as the electrically evoked compound action potential (eCAP), to supplement behavioral information. We investigated whether eCAPs could be recorded in patients with a vestibulo-cochlear implant. Specifically, eCAPs were successfully recorded for cochlear and vestibular setups, as well as for mixed cochlear-vestibular setups. Similarities and slight differences were found for the recordings of the three setups. These findings demonstrated the feasibility of eCAP recording with a vestibulo-cochlear implant. They could be used in the short term to reduce current spread and avoid activation of non-targeted neurons. More research is warranted to better understand the neural origin of vestibular eCAPs and to utilize them for clinical applications.

Keywords: bilateral vestibular loss; cochlear implant; electrically evoked compound action potential; neural prosthesis; vestibular function; vestibular implant; vestibular prosthesis.

Figures

Figure 1

Amplitude growth functions for the cochlear setup in subjects 1–4. The gray inset in the top left corner of every panel illustrate the eCAP response to the highest current amplitude in that subject. The AGFs were measured for 12 different current amplitudes in each subject and were recorded with apical electrode contacts in subjects 1–3 and basal electrodes for subject 4. A clear N-P complex was best visible in the eCAP response of subject 1, while the slope of the AGF was largest in subject 3 (2.3 μV/cu).

Figure 2

Amplitude growth functions for the mixed setup in subjects 1–4. The gray insets in the top left corner of each panel represent the eCAP response to the largest current amplitude. eCAPs and AGFs were measured with a stimulating electrode in the cochlea and a recording electrode in a semicircular canal (Top row) and mirrored conditions (Bottom row). A clear N-P complex was only observed in subjects 3 and 4 (S8R11 and S8R10, respectively). Slopes of the AGFs were generally smaller than for the cochlear setup (between 0.1 and 0.8 μV/cu for mixed setup, between 1.0 and 2.3 μV/cu for cochlear setup). S10R8 for subject 4 was grayed out here and in following plots as the eCAP response was not likely to be a neuronal response.

Figure 3

Amplitude growth functions for the trans-canal setup in subjects 1–4. The gray insets in the top left corner of each panel represent the eCAP response to the largest current amplitude. eCAPs and AGFs were measured with a stimulation electrode in one semicircular canal and the recording electrode in another canal. A clear N-P complex could be observed only in subjects 1 and 3. Subject 3 had the largest N-P magnitudes for this setup and the corresponding slope was 4.1 μV/cu. In the case of subject 1, the N-P complex was less pronounced and the AGF slope was 0.6 μV/cu. eCAPs and AGFs for both subjects 2 and 4 were grayed out from here, as the responses were unlikely of neuronal origin, but rather a result of stimulation artifact.

Figure 4

Magnitudes and latencies of eCAPs in all setups. For a given setup, the box captures the range of N-P magnitudes (width of box) and the range of N-P latencies (height of box). The bottom of the box represents the earliest N peak, while the top of the box represents the latest P peak. These values were taken from the averaged eCAP responses across all current amplitudes (cf. Figures 1–3). Latencies of N-P peaks were comparable across setups and patients. N peaks occurred around 250 μs after stimulation onset and P peaks occurred between 600 and 850 μs. N-P magnitudes denoted by the width of the black boxes were similar for cochlear and trans-canal setups, whereas magnitudes for mixed setup were smaller (cf. subjects 1 and 3). Pairs for subjects 2 and 4 were grayed out as they were likely the result of stimulation artifact.

Figure 5

Distances between stimulation, recording electrodes and ampullae illustrated as triangles. A black dot denotes the stimulation electrode, a red dot the recording electrode, and a blue dot the ampulla of the corresponding vestibular electrode (e.g., the lateral ampulla for electrode 10). For cochlear setups, ampullae were not applicable, whereas for trans-canal setups the ampulla to the corresponding recording vestibular electrode was used (e.g., the superior ampulla for a recording electrode 11). The text above each triangle notes the electrode pair tested. The values are the distances in millimeters. First, the distance between the stimulation and recording electrodes; second, between the stimulation electrode and the ampulla; and third, between the recording electrode and ampulla.

Figure 6

Distances between stimulation, recording electrodes and ampullae vs. maximum N-P voltage (A) and N-P latencies (B,C, respectively). The different setups are represented through different markers. No clear correlation could be established between distances and N-P voltages or latencies.

Figure 7

Influence of impedance on the maximum N-P voltage. No general impact could be inferred from the recordings. Gray data points represent discarded sets of subjects 2 and 4 that were likely influenced by stimulation artifact.

Figure 8

eCAP responses during continuous stimulation of 50 min in subject 1 (experiment B1). (A) eCAP was recorded at 10 min intervals in response to a pulse with the same current amplitude as the continuous stimulation (but a shorter phase-width, 50 vs. 200 μs, average of 60 iterations). Only at 0.8 min did the eCAP response have a similar morphology to that observed in experiment A without continuous stimulation (gray inset at 0.8 min for comparison). At that time point the N peak was still clearly visible, but for later time points only a part of it was visible. (B) The N-P amplitude measured in this experiment was smaller than that measured experiment A without continuous stimulation at all time points, the eCAP responses here had a smaller N-P amplitude. The gray horizontal line was the N-P magnitude from experiment A for this current amplitude.

Figure 9

eCAP responses during continuous stimulation of 50 min in subject 1 (experiment B2). (A) eCAP was recorded in response to a pulse with the maximal current amplitude at 10 min intervals (average of 60 iterations). No significant difference in morphology was observed with increasing duration of stimulation (gray inset at 0.8 min from experiment A for comparison). (B) N-P magnitudes were smaller at all time points than for the same pulse without continuous stimulation in experiment A (gray horizontal line).