Electrically evoked auditory steady state response detection in cochlear implant recipients using a system identification approach

Regular Cochlear implant (CI) fitting is an important aspect of hearing restoration of CI recipients. CI stimulation parameters such as the so-called T- and C-levels are tuned to the recipient's needs and are crucial in hearing performance. Electrically evoked auditory steady state responses (EASSRs) are neural responses elicited by amplitude modulated stimulation pulse trains. They can be measured via electroencephalography (EEG) and have shown to be possible objective measures for CI fitting, which could lead to an automation of the fitting procedure and hence improve clinical care. However, artifacts in the EEG recording, originating from the CI stimulation, hamper EASSR detection. A number of stimulation artifact removal methods have been introduced and applied with different levels of success. EEG recordings with clinically relevant stimulation parameters, especially from ipsilateral recording channels, remain difficult to analyze. In this paper, we present a novel approach, which models EASSR and stimulation artifact using a system identification procedure. We use the apparent latency to compare its performance with that of the benchmark approach based on linear interpolation. The observed results suggest better stimulation artifact removal and EASSR detection with the new approach, especially for ipsilateral EEG recording channels.