CI in UHL low-frequency pitch perception (Dillon et al., 2019)

Purpose: Three experiments were carried out to evaluate the low-frequency pitch perception of adults with unilateral hearing loss who received a cochlear implant (CI).
Method: Participants were recruited from a cohort of CI users with unilateral hearing loss and normal hearing in the contralateral ear. First, low-frequency pitch perception was assessed for the 5 most apical electrodes at 1, 3, 6, and 12 months after CI activation using an adaptive pitch-matching task. Participants listened with a coding strategy that presents low-frequency temporal fine structure (TFS) and compared the pitch to that of an acoustic target presented to the normal hearing ear. Next, participants listened with an envelope-only, continuous interleaved sampling strategy. Pitch perception was compared between coding strategies to assess the influence of TFS cues on low-frequency pitch perception. Finally, participants completed a vocal pitch-matching task to corroborate the results obtained with the adaptive pitch-matching task.
Results: Pitch matches roughly corresponded to electrode center frequencies (CFs) in the CI map. Adaptive pitch matches exceeded the CF for the most apical electrode, an effect that was larger for continuous interleaved sampling than TFS. Vocal pitch matches were variable but correlated with the CF of the 3 most apical electrodes. There was no evidence that pitch matches changed between the 1- and 12-month intervals.
Conclusions: Relatively accurate and asymptotic pitch perception was observed at the 1-month interval, indicating either very rapid acclimatization or the provision of familiar place and rate cues. Early availability of appropriate pitch cues could have played a role in the early improvements in localization and masked speech recognition previously observed in this cohort.

Supplemental Material S1. The association between performance on individual sound field measures (y-axis) and the normalized mean pitch match for each electrode (x-axis). The top row displays the speech recognition with the CI-alone (masking to the NH-ear) as measured with CNC words in quiet. Performance is scored as the percent correct. The middle rows display the speech recognition in noise with the CI plus the NH-ear (CI+NH) as measured with the AzBio sentences in noise. Performance is scored as the percent correct. The AzBio sentences were presented from directly in front of the listener, with the noise presented either co-located with the signal (S0N0) or spatially separated to the side contralateral from the CI (S0NContra). The bottom row of panels shows RMS error for sound source localization in the CI+NH condition, where a lower value indicates better performance.

Dillon, M. T., Buss, E., Rooth, M. A., King, E. R., Pillsbury, H. C. & Brown, K. D. (2019). Low-frequency pitch perception in cochlear implant recipients with normal hearing in the contralateral ear. Journal of Speech, Language, and Hearing Research, 62, 2860–2871. https://doi.org/10.1044/2019_JSLHR-H-18-0409