Mismatch Negativity (MMN) in Response to Noise-Suppressed Speech as Perceived by Cochlear Implant Users - OMAI 2018- The 10th International Symposium on Objective Measures in Auditory Implants

Chin-Tuan Tan ^2,3 Fang Yu ¹ Fang Yu ¹

¹Department of Electrical and Electronic Engineering, Southern University of Science and Technology, Shenzhen, China
²Department of Electrical Engineering,Erik Jonsson School of Engineering and Computer Science, The University of Texas at Dallas, Dallas, USA
³School of Behavioral and Brain Science, The University of Texas at Dallas, Dallas, USA

Background: Speech perception in noise is always a challenge to cochlear implant (CI) users, and noise-suppression processing is commonly utilized to enhance speech for better perception. An objective metric to measure the effectiveness of such processing will be useful and necessary.

Objective: To investigate mismatch negativity (MMN) response to noise-suppressed speech as perceived by CI-users and normal hearing (NH) listeners.

Methods: 2 CI users and 14 NH listeners participated in this study. They are all Mandarin speakers. A steady-state noise was added to a vowel /a/ at signal-to-noise ratio (SNR) of -5 and +5 dB, and were then separately noise suppressed. EEG recording was divided into two sessions. Two noisy stimuli were presented as the deviant stimuli in one session and two noise-suppressed stimuli as the deviant stimuli in second session, keeping the clean vowel as the standard stimulus. All stimuli were presented at 65 dB SPL in a sound treated room.

Preliminary results: Experimental results with CI users showed shorter MMN latencies with the noise-suppressed stimuli than with the noisy stimuli, which is similar to the pattern observed with NH participants. However, the noise-suppressed stimuli evoked larger MMN amplitudes as compared to those elicited by the noisy stimuli, which was not found in NH participants.

Conclusion: Currently, MMN is responding to the effect of noise-suppression on speech perception by NH listeners and CI users in a different manner. More data will be necessary to establish MMN as a potential correlate to noise-suppressed speech by CI users.