Channel crosstalk detected using ECAP measurements is associated with poorer speech perception in cochlear implant users

• Channel independence was assessed with electrically evoked compound action potentials. • In some cases, distant maskers were as or more effective than same-electrode maskers. • Substantial atypical masking or ‘crosstalk’ was seen in thirteen of thirty adult CI users. • The large majority of crosstalk subjects had poor sentence recognition in noise. • Channel crosstalk detected via ECAPs may be a biomarker for poor neural survival. The number and independence of channels in cochlear implants (CI) has long been considered to influence speech recognition, particularly in competing background noise. Measures of channel independence have been obtained via psychophysical and objective means, relying on interactions between probe and masker signals delivered on different channels. In the current study, electrically evoked compound action potentials (ECAP) obtained from 32 Nucleus CI recipients tested at one basal and one apical position were performed using a standard spread-of-excitation procedure. An alternative analysis method, comparing masked responses only, revealed distant maskers as effective or more effective than same-electrode maskers in 13/32 cases. This appears to indicate substantial crosstalk between channels, covering up to nine intracochlear electrodes in one subject. Subjects with atypical responses and no other limiting factors had significantly poorer sentence recognition in noise compared with those with no detected peripheral or cognitive limiting factors. We propose that channel crosstalk detected via ECAPs may be a biomarker for poor or patchy neural survival that leads to poorer speech perception in CI recipients.