Abstract
Background
Pitch perception is an essential element of human communication, which is hindered in cochlear implant (CI) users. Past studies with more artificial stimuli, such as single-channel pulse trains, have suggested that pitch perception via a CI relies primarily on temporal cues as opposed to via normal hearing (NH) that relies heavily on spectral cues. However, recent work involving more realistic stimuli challenges this separation of mechanisms between NH and CI listeners. The latter studies suggest that implant users might also be utilizing spectral cues in addition to the temporal cues. In the present study, we aim to clarify the relative contributions of spectral and temporal cues to pitch perception of CI users using realistic speech stimuli, such as syllables, words, and short sentences.
Methods
A set of stimuli will be generated by manipulating the F0 of Dutch syllables, words, and short sentences with the STRAIGHT program in MATLAB. We will assess the prevalence of both cues in the stimuli by recording the output of CI processors, using the implant-in-a-box’s from Cochlear Ltd and Advanced Bionics. The pitch cues of these recordings will then be manipulated to remove temporal cues and/or spectral cues, creating four types of speech stimuli: unaltered, or spectral, temporal, or both cues removed. Using a research interface, the F0 pitch discrimination abilities of CI participants will be assessed with adaptive alternative-forced choice (AFC) tasks which yield F0 just-noticeable differences (JNDs) and psychometric functions.
Results and Conclusions
With this method we will be able to systematically inspect the role of both cues across parameters concerning the nature of the stimuli and implant stimulation strategy. Our expectations for the results are: 1.) F0 JNDs will increase with decreasing pitch cue availability. 2.) When both spectral and temporal cues are removed, CI users will be unable to converge to a JND. 3.) The relative contributions of spectral and temporal cues will depend on the content of the stimulus as well as the stimulation strategy. The findings will clarify which type of cue should be enhanced to improve pitch perception of CI users, depending on the nature of the sound and the coding strategy.
Pitch perception is an essential element of human communication, which is hindered in cochlear implant (CI) users. Past studies with more artificial stimuli, such as single-channel pulse trains, have suggested that pitch perception via a CI relies primarily on temporal cues as opposed to via normal hearing (NH) that relies heavily on spectral cues. However, recent work involving more realistic stimuli challenges this separation of mechanisms between NH and CI listeners. The latter studies suggest that implant users might also be utilizing spectral cues in addition to the temporal cues. In the present study, we aim to clarify the relative contributions of spectral and temporal cues to pitch perception of CI users using realistic speech stimuli, such as syllables, words, and short sentences.
Methods
A set of stimuli will be generated by manipulating the F0 of Dutch syllables, words, and short sentences with the STRAIGHT program in MATLAB. We will assess the prevalence of both cues in the stimuli by recording the output of CI processors, using the implant-in-a-box’s from Cochlear Ltd and Advanced Bionics. The pitch cues of these recordings will then be manipulated to remove temporal cues and/or spectral cues, creating four types of speech stimuli: unaltered, or spectral, temporal, or both cues removed. Using a research interface, the F0 pitch discrimination abilities of CI participants will be assessed with adaptive alternative-forced choice (AFC) tasks which yield F0 just-noticeable differences (JNDs) and psychometric functions.
Results and Conclusions
With this method we will be able to systematically inspect the role of both cues across parameters concerning the nature of the stimuli and implant stimulation strategy. Our expectations for the results are: 1.) F0 JNDs will increase with decreasing pitch cue availability. 2.) When both spectral and temporal cues are removed, CI users will be unable to converge to a JND. 3.) The relative contributions of spectral and temporal cues will depend on the content of the stimulus as well as the stimulation strategy. The findings will clarify which type of cue should be enhanced to improve pitch perception of CI users, depending on the nature of the sound and the coding strategy.
| Original language | English |
|---|---|
| Publication status | Published - 29-Jun-2023 |
| Event | 4th Virtual Conference on Computational Audiology - Duration: 29-Jun-2023 → 30-Jun-2023 https://computationalaudiology.com/events/vcca2023/ |
Conference
| Conference | 4th Virtual Conference on Computational Audiology |
|---|---|
| Abbreviated title | VCCA2023 |
| Period | 29/06/2023 → 30/06/2023 |
| Internet address |