Association between ECAP signal artifacts and electrical components from a cochlear implant impedance model

electrically evoked compound action potentials (ECAPs) are among the few objective measures of auditory nerve activity available in clinical practice and provide valuable insight into a patient's neural status. ECAP recordings are typically performed using different electrodes for stimulation and recording to avoid amplifier saturation caused by electrode polarization. However, same-electrode recording could reduce the spatial separation between the site of neural activation and the recording site, potentially providing a more localized assessment of neural activity and improving the sensitivity of ECAP-based biomarkers. Here, we introduce a method to characterize stimulation artifacts and estimate their parameters using a previously published impedance model. The artifact waveform was described by a shared double exponential function, while electrical components of an equivalent circuit were estimated from impedance measurements acquired with multiple pulse durations. Artifact recordings exhibited clear polarity-dependent behavior. Nevertheless, the amplitude of the fast exponential component showed a significant association with its impedance-based estimate (fixed-effect coefficient β=0.37 and β=0.44 for anodic-cathodic and cathodic-anodic stimulation, respectively; p<0.001). Furthermore, longitudinal changes in the fast exponential component were similar to changes in near-field impedance subcomponent, suggesting that the ECAP artifact could be partially influenced by postoperative changes at the electrode-electrolyte interface.