Effects of stimulus polarity on the local evoked potential in auditory brainstem implant users

Auditory brainstem implants (ABI) can enable hearing sensation through electrical stimulation of the cochlear nucleus. The basic stimulation and signal coding strategies of the ABI are based on those of the cochlear implant. This may not always be optimal, and ABI-specific strategies may be preferred. In a cohort of ten ABI users, we examined the feasibility of measuring local evoked potentials (LEP) via fine-grained stimulation with a forward masking paradigm. We introduce a new baseline-dependent definition of LEP amplitude for analyzing the LEP amplitude growth function to obtain threshold stimulation levels and slope values. The processing of biphasic pulses by the cochlear nucleus and the influence of the leading phase polarity were examined. There were no statistically significant differences in LEP thresholds or slopes between cathodic and anodic leading pulses. LEP thresholds measured with cathodic leading pulses (r = 0.77, t₃₁ = 6.81, p < 0.0001) and anodic leading pulses (r = 0.70, t₂₇ = 45.14, p < 0.0001) correlated significantly with perceptual hearing thresholds. The correlation analysis was impacted by outlier values, especially in the case of LEP thresholds measured with anodic leading pulses. Cathodic leading pulses had significantly shorter LEP peak latencies (t_104.8 = 2.63, p < 0.01). These results show that the cathodic leading pulses are superior for eliciting LEPs. We suggest that cathodic leading pulses should be the basis for ABI-specific coding strategies.