2019 Conference on Implantable Auditory Prostheses

14-19 July 2019

Granlibakken, Lake Tahoe

Page 147

M56: EVALUATING INDIVIDUALIZED COCHLEAR IMPLANT PROCESSORS WHOSE SITES

OF STIMULATION WERE SELECTED USING THE INTERPHASE GAP EFFECT AND A

COMBINATION OF PHYSICAL AND VIRTUAL ELECTRODES

Joshua S. Stohl, Kostas Kokkinakis

MED-EL Corporation, Durham, NC, USA

One hypothesized source of variability in outcomes with cochlear implants (CIs) is the health of the

auditory periphery. Evoked compound actional potentials (ECAPs) can be used to make direct

measurements of peripheral neural responses, and animal model studies have shown that the ECAP-

based interphase gap (IPG) effect, which is the difference between the slopes of two ECAP amplitude

growth functions (AGFs) measured in response to waveforms with two different IPGs, is highly correlated

with spiral ganglion cell packing density (e.g., Ramekers et al., 2014, J Assoc Res Otolaryngol, 15, 187-

202). It has also been shown that, for bilateral users, the between-ear difference in the across-site mean

IPG effect is correlated with the between-ear difference in speech recognition measures (Schvartz-

Leyzac et al., 2018, Ear Hear, 39, 344-358). Thus, we sought to test the hypothesis that increasing the

mean IPG effect within single ears would improve speech recognition abilities. Additionally, rather than

limit site selection to physical electrodes, the IPG effect was measured for both physical and virtual

electrodes, where virtual electrodes corresponded to parallel stimulation by two adjacent physical

electrodes.

In the first part of this study, virtual electrodes were compared to physical electrodes using ECAP

waveforms and compound discharge latency distributions, AGF slopes, and the IPG effect. Although

several investigators have previously compared virtual electrodes to physical electrodes using ECAP-

based measures, such as spread of excitation, to our knowledge, this study is the first to do so in

patients implanted with CIs manufactured by MED-EL, and to assess AGF slopes and IPG effect for

virtual sites. The difference in device manufacturer is noteworthy because electrode arrays made by

MED-EL typically (1) extend to more apical regions of the cochlea, and (2) have a wider electrode

spacing compared to previously studied devices.

In the second part of this study, the IPG effect was measured at every electrode. Processors

were then created with the aim of using the IPG effect for controlling stimulation. For all processors, the

most apical and basal electrodes were always fixed to preserve the total stimulated cochlear region.

Two different versions of each processor were created, one in which the frequency allocation of

the filter bank was fixed across processing conditions and equivalent to MED-

EL’s clinical, modified

logarithmic distribution of center frequencies (Logarithmic FS), and a second in which the filter bank

center frequencies were adjusted based on the selected sites of stimulation. The latter set of processors

was implemented to mitigate the potential confound of changes to the place-frequency map (due to

changes in electrode assignments for each frequency band) during the acute behavioral evaluation of

the processors.

Subjects performed monosyllabic words tests and sentence tests in quiet and in noise, with

stimuli delivered via the MAX clinical interface controlled by software developed for the Research

Interface Box, version 2 (RIB2) and custom research software. ECAP measures and speech recognition

data obtained with the individually-optimized processors will be presented.