01whole.pdf (5.57 MB)
Exploring place pitch & temporal pitch perception with cochlear implants
thesisposted on 2022-03-28, 15:02 authored by Vijay Mohan Raaja Marimuthu
Cochlear Implant (CI) recipients are on par with their normal hearing counterparts in speech perception tasks in quiet, but their performance is at chance level for pitch perception tasks. Pitch in CI can be conveyed by either place cues or temporal cues. Although researchers have studied place and temporal pitch, there is scant evidence that reports the individual contribution of place and temporal pitch as a function of musical pitch. This thesis aims to investigate the role of place pitch and temporal pitch perception in CI recipients. It comprises three major experiments using four experimental procedures. The four experimental procedures were: 4AFC Discrimination, 2AFC Ranking, and 2AFC Modified Melodies Test – Backward, and 2AFC Modified Melodies Test – Warp modification.Experiment I investigated temporal pitch sensitivity as a function of different stimulation patterns, base pulse rate, and electrode location. The four stimulation patterns were: single electrode stimulation (apical (Electrode E22) or middle (E12)), dual-electrode stimulation (E22 & E12), and multiple electrode stimulation (E22 to E12). Six post-lingually deafened CI subjects were tested using three of the four experimental procedures (i.e., excluding the discrimination task). The stimuli were presented as pulses at a base rate of 131 pulses per second (pps) (musical note C3 range) and 262 pps (C4 range). The temporal pitch sensitivity was not influenced by different stimulation patterns and the results showed no significant difference among the stimulation patterns across the three procedures. The results suggest that the CI recipients are unable to combine cues from different places in the cochlea to give a “stronger” cue.Experiment II aimed to investigate the individual contribution of place pitch and temporal pitch in various pitch perception tasks. The performance of CI subjects was tested using four experimental procedures and three stimulus types. The three stimulus types were: (1) Pure tones with base frequency of C5 (523 Hz), providing place cues only; (2) Harmonic tones with base frequency of C3 (131 Hz), providing temporal cues only; (3) Harmonic tones with base frequency of C4 (262 Hz), providing both place and temporal cues. The stimuli were presented via loudspeaker at a comfortable loudness level. The recipients used their own speech processor. The overall scores for discrimination and ranking were high for all three stimulus types, however, three subjects showed pitch reversals in ranking the C4 harmonic tones. In the Modified Melodies test, scores were similar for C5 pure tones and C3 harmonic tones, while scores using C4 harmonic tones were worse and mostly near chance. These results suggest that CI place pitch may convey melodic pitch information, but the contribution of brightness cannot be completely ruled out.Experiment III sought to investigate the role of brightness in various pitch perception tasks in normal hearing individuals. The goal of this study was to investigate whether CI subjects in the previous experiment perceived place pitch as pitch rather than as a pattern of brightness changes. Eighteen normal-hearing adults participated in four experimental procedures using three stimulus types: brightness sequences (harmonic tones varying in brightness, with constant pitch), and noise sequences (Low-pass noise bands, varying in cut-off frequency) were compared with pitch sequences (harmonic tones varying in pitch, with constant brightness). Results showed that the subjects were able to discriminate and rank brightness, and were able to detect brightness contour changes, but were unable to make judgements of musical intervals for brightness. These results suggest that the cochlear implant recipients in the previous experiment may have perceived place cues as brightness rather than pitch.In conclusion, a similar performance trend was seen between CI place pitch and NH brightness indicating that CI place pitch is akin to brightness and not to pitch. Conversely, CI performance was similar between temporal pitch and place pitch suggesting that CI place pitch can convey melodic pitch. However, the overall performance reveals that CI place pitch is more akin to brightness aspect of timbre than to pitch and additional studies need to affirm these findings. Unfortunately, the CI pitch performance was still significantly below NH performance.