Decoding Selective Attention in Normal Hearing Listeners and Bilateral Cochlear Implant Users With Concealed Ear EEG.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Auditory Perception

Background:

• Electroencephalography (EEG) can decode attended speech in normal-hearing (NH) listeners.
• Applications include identifying target speakers and guiding cochlear implant (CI) speech enhancement.
• CI-induced artifacts and lower spectral resolution may hinder EEG applicability in CI users.

Purpose of the Study:

• Investigate selective attention decoding in CI users using an around-the-ear EEG system (cEEGrid).
• Compare cEEGrid performance against high-density cap EEG for attention decoding.

Main Methods:

• A selective attention paradigm with two concurrent audiobooks was used.
• Simultaneous EEG recordings via a 96-channel cap and an 18-channel cEEGrid.
• Envelope tracking algorithm evaluated reconstruction performance against attended/unattended speech envelopes.

Main Results:

• High-density EEG successfully decoded selective attention in NH and CI users.
• 9 out of 10 CI users and all NH listeners achieved high decoding accuracies.
• cEEGrid decoded selective attention in 50% of NH listeners and 50% of CI users.

Conclusions:

• Decoding selective attention using single-trial EEG data is feasible in both NH and CI users.
• The cEEGrid system shows potential for attention decoding in CI users, though performance may vary.
• Further research could optimize cEEGrid for improved CI user auditory attention decoding.