Temporal fine structure processing, pitch and speech perception in cochlear implant recipients

Cochlear implant (CI) recipients usually complain about poor speech understanding in the presence of noise. Indeed, they generally show ceiling effects for understanding sentences presented in quiet, but their scores decrease drastically when testing in the presence of competing noise. One important aspect that contributes to speech perception skills, especially when listening in a fluctuating background, has been described as Temporal Fine Structure (TFS) processing. TFS cues are more dominant in conveying Low Frequency (LF) signals linked in particular to Fundamental Frequency (F0), which is crucial for linguistic and musical perception. A§E Harmonic Intonation (HI) and Disharmonic Intonation (DI) are tests of pitch perception in the LF domain and their outcomes are believed to depend on the availability of TFS cues. Previous findings indicated that the DI test provided more differential LF pitch perception outcomes in that it reflected phase locking and TFS processing capacities of the ear, whereas the HI test provided information on its place coding capacity as well. Previous HI/DI studies were mainly done in adult population showing abnormal pitch perception outcomes in CI recipients and there was no or limited data in paediatric population as well as HI/DI outcomes in relation to speech perception outcomes in the presence of noise. One of the primary objectives of this thesis has been to investigate LF pitch perception skills in a group of pediatric CI recipients in comparison to normal hearing (NH) children. Another objective was to introduce a new assessment tool, the Italian STARR test which was based on measurement of speech perception using a roving-level adaptive method where the presentation level of both speech and noise signals varied across sentences. The STARR test attempts to reflect a better representation of real world listening conditions where background noise is usually present and speech intensity varies according to vocal capacity as well as the distance of the speaker. The Italian STARR outcomes in NH adults were studied to produce normative data, as well as to evaluate interlist variability and learning effects. Finally, LF pitch perception outcomes linked to availability of TFS were investigated in a group of adult CI recipients including bimodal users in relation to speech perception, in particular Italian STARR outcomes. Results were interesting: Although the majority of CI recipient children showed abnormal outcomes for A§E, their scores were considerably better than in the adult CI users. Age had a statistically significant effect on performance in both children and adults; younger children and older adults tended to show poorer performance. Similarly, CI recipient adults (even the better performers) showed abnormal STARR outcomes in comparison to NH subjects and group differences were statistically significant. The duration of profound deafness before implantation had a significant effect on STARR performance. On the other hand, the significant effect of CI thresholds re-emphasized the sensitivity of the test to lower level speech which a CI user can face very often during everyday life. Analysis revealed statistically significant correlations between HI/DI and STARR performance. Moreover, contralateral hearing aid users showed significant bimodal benefit for both HI/DI and STARR tests. Overall findings confirmed the usefulness of evaluating both LF pitch and speech perception in order to track changes in TFS sensitivity for CI recipients over time and across different listening conditions which might be provided by future technological advances as well as to study individual differences.