Templates for speech-evoked auditory brainstem response performance in cochlear implantees

Abstract

Introduction Speech-evoked auditory brainstem response (ABR) has been used to assess the fidelity of encoding speech stimuli at the subcortical level in normal individuals in noise and in special populations such as learning-impaired children and musicians. The neural code generated by cochlear implants (CIs) in the auditory brainstem pathway and its similarity to stimulus may account for variable speech development in cochlear implantees. Objective The aim of this study was to describe speech ABR recorded in CI individuals and establish measurement parameters for the neural response and its reproducibility. Participants and methods Children between 5 and 10 years of age implanted in the right ear with fully inserted 12-electrode CIs were selected. All participants had normal morphology of the cochlea and auditory nerve in preoperative computed tomographic scan and MRI. Speech syllable 40 ms /da/ was used to elicit speech ABR. Response traces for intensity input/output functions were harvested. Grand averages were constructed for peak picking. Individual patient responses were analyzed for reproducibility, latency of wave V, root mean square amplitude of the response, and correlation to the stimulus. Results Grand averages showed wave V, followed by the frequency following response. Wave V is a vertex-positive peak, equivalent to that elicited by a click, which reflects the stimulation by the transient /d/. The mean latency of wave V was 2.59±0.7 ms at 70 dBHL. The frequency following response showed multiple sequenced troughs corresponding to the sustained vowel /a/. Individual responses collected for similar stimulus parameters showed high reproducibility, being 99.65% at 60 dBHL and 52.8% at 30 dBHL. Participants showed variable latency and root mean square amplitude-intensity input–output functions slopes. The mean stimulus-to-response correlation was 18.1±3.1%. Conclusion Speech ABR in CI participants shows similar morphology to that recorded in norms. CIs thus transcribe the speech signal with high fidelity to the brainstem pathways.