Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Influence of Model Voice Cues and Hearing Impairment on Behavioural and Cortical Responses to the Pitch Shift Reflex
Poster Session E, Friday, October 2, 11:00 am - 1:00 pm, Wangari Maathai
Tony Schelhorn1,2, Benjamin Peschel2, Martijn Wieling1, Ulrich Hoppe2, Michael Döllinger2, Defne Abur1; 1Rijksuniversiteit Groningen, 2Universitätsklinikum Erlangen
Background: Altered auditory feedback (AAF) tasks can grant insights into neural control of speech, including how auditory perceptual information influences speech motor production. In AAF tasks, participants are asked to produce speech while their auditory feedback (via headphones) is shifted in realtime (e.g., the pitch is increased). Speakers typically respond by opposing the shift (e.g., lowering their own vocal pitch) but there is high variability in responses. To reduce vocal variability, model voice cues can be used, wherein speakers hear a model voice before each trial. However, it is not clear how the distance between a speaker’s habitual pitch and the pitch of the model voice (’pitch distance’) impacts responses. The current study assessed how pitch distance influenced behavioural and cortical responses as well as voice stability during an AAF experiment. Given that this relationship may vary with the degree of auditory feedback, the task was evaluated in speakers with differing hearing status. Methods: Participants with typical hearing (N = 20; 13 female, 7 male, median age: 61 years) and with bilateral sensory-neural hearing impairment of ≥ 40 dBHL (N = 15; 5 female, 10 male, median age: 76 years) completed the study. All participants completed written consent in accordance with the ethical review board of the Friedrich-Alexander-Universität Erlangen-Nürnberg (no.: 423_20 B). Habitual pitch was calculated by averaging fundamental frequency across two speech tasks (counting from 20 to 30, and producing the vowel /a/). Voice stability parameters were calculated from 40 baseline (no shift) productions of the vowel /a/ during the experiment. In the AAF experiment, a model voice cue (/a/) was presented before each trial (with a pitch of 220 Hz for female and 110 Hz for male) and participants were instructed to match it. Between 1 - 1.3 seconds later, a +2 semitone pitch shift was applied in the headphone feedback. This was repeated for 40 trials. Responses were measured behaviourally (fundamental frequency via microphone) and cortically (N1 and P2 responses via EEG). Linear mixed effects models were used to analyse the impact of pitch distance, speaker group, and sex on behavioural and cortical responses and voice stability parameters. Results: Pitch distance had a significant effect across observed parameters for the typical hearing group (p < 0.001), but not the hearing impaired group (p = 0.822). In both groups, female participants had lower habitual pitches (mean: 180.5 Hz) compared to the female model voice cue. Conclusion: Model voice cues can reduce voice variability for AAF experiments; however, the perceptual pitch distance can influence responses and voice stability. Future work should directly compare responses with and without a model voice to determine comparability of AAF findings using voice cues.
Topic Areas: Speech Motor Control,