Characterizing Cochlear implant artefact removal from EEG recordings using a real human model

Area of Science:

• Neuroscience
• Biomedical Engineering
• Auditory Neuroscience

Background:

• Electroencephalography (EEG) recordings in cochlear implant (CI) users are often obscured by electrical artefacts.
• Existing methods like interpolation and spatial filtering have been used, but their effectiveness requires further investigation.

Purpose of the Study:

• To assess the efficacy of interpolation and spatial filtering in removing electrical artefacts from EEG recordings in CI listeners.
• To determine the optimal conditions for artefact reduction across different stimulation parameters.

Main Methods:

• Simulated electrical artefacts were generated using amplitude-modulated pulse trains at various rates (100-902 pps) in a human head specimen with a CI.
• Template waveforms for auditory change complex (ACC), auditory steady-state response (ASSR), and auditory change following response (AC-FR) were added to contaminated recordings.
• The effectiveness of interpolation and spatial filtering was evaluated based on artefact reduction and waveform distortion.

Main Results:

• Interpolation effectively removed artefacts for pulse rates below 400 pps.
• Both interpolation and spatial filtering were effective at higher pulse rates (>400 pps).
• Minimal distortions were observed for ACC and AC-FR, while ASSR showed some amplitude and phase distortions.

Conclusions:

• Interpolation and spatial filtering are effective techniques for reducing electrical artefacts in CI listener EEG data.
• These methods allow for the recovery of reliable neural responses, even at high stimulation rates.
• The findings support the use of these artefact removal strategies in clinical and research settings involving CI users.