Presentation

Alexis Basciano¹, Corianne Rogalsky¹, Ayoub Daliri¹, Madilyn Pettijohn¹, Madilyn Majors¹; ¹Arizona State University

Sensorimotor integration is crucial for intelligible speech and to correct mistakes during speech production. Delayed auditory feedback (DAF), a paradigm used to test sensorimotor integration in speech, delays auditory feedback during speech production in real time to examine online changes in speech fluency. DAF causes disfluencies in neurotypical populations but increases fluencies in individuals who stutter (Daliri et.al, 2018). Because sensorimotor integration is impaired in post-stroke aphasia (Buchsbaum et.al, 2011), DAF can be used to characterize speech production impairments in aphasia. The first DAF study in aphasia found that a delay was related to decreased dysfluencies in conduction aphasia (n=3), while individuals with nonfluent aphasia performed similar to controls, i.e. more dysfluencies under DAF (Boller et. al, 1978). More recent work indicates that perturbing the auditory feedback of speech may be a promising avenue for understanding the mechanisms of speech production impairments in aphasia (Behroozmand et. al, 2018). Thus, the purpose of the current study is to revisit delayed auditory feedback to characterize speech production impairments in individuals with aphasia, using quantitative acoustic measures. To date, we have recruited six individuals with aphasia and nine age-matched controls to complete a sentence-reading task using DAF at 4 different delay intervals (0 ms, 100 ms, 200 ms, and 400 ms). Five target sentences were used, which were each 8 syllables in length and used American-English phonemes. Speech production was characterized using linguistic categories (content errors, morphological errors, phonological errors, and dysfluencies for each utterance) and acoustic features (F1, F2, amplitude, and duration for each type of phoneme – vowels, diphthongs, voiced consonants, unvoiced consonants). ANOVAs and single-case Bayesian statistics were used to characterize the performance of the individuals with aphasia as a function of delay duration. Preliminary analyses indicate that, compared to controls, the aphasia group generated more errors in content word, morphological, and phonological categories across all delay intervals; there was no significant main effect of delay or group by delay interaction. However, the acoustic analyses and single case statistics indicate substantial variability across the individuals with aphasia. For example, under the normal speaking condition with no delay, the participant with conduction aphasia exhibited significantly larger vowel amplitudes and longer vowel durations than controls; however, with increasing delay, vowel durations became not significantly different than controls. In other words, without delay, the conduction aphasia participant’s speech was louder and longer compared to controls but became more control-like with increased delays (I.e., DAF improved or normalized the participant’s speech). Perhaps longer delay periods allowed the individual with conduction aphasia to engage in compensatory strategies that remedied their abnormal vowel durations, but not their amplitude. Overall, the preliminary results of this study indicate DAF may affect the acoustic, but not linguistic, properties of speech production in individuals with aphasia. Perhaps after more studies are conducted, DAF can be used to characterize speech impairments in aphasia and in the future serve as a tool to improve their speech.

Topic Areas: Multisensory or Sensorimotor Integration, Speech Motor Control