Journal of Speech Language and Hearing Research最新文献

Purpose: Phonetic forced alignment has a multitude of applications in automated analysis of speech, particularly in studying nonstandard speech such as children's speech. Manual alignment is tedious but serves as the gold standard for clinical-grade alignment. Current tools do not support direct training on manual alignments. Thus, a trainable speaker adaptive phonetic forced alignment system, Wav2TextGrid, was developed for children's speech. The source code for the method is publicly available along with a graphical user interface at https://github.com/pkadambi/Wav2TextGrid.

Method: We propose a trainable, speaker-adaptive, neural forced aligner developed using a corpus of 42 neurotypical children from 3 to 6 years of age. Evaluation on both child speech and on the TIMIT corpus was performed to demonstrate aligner performance across age and dialectal variations.

Results: The trainable alignment tool markedly improved accuracy over baseline for several alignment quality metrics, for all phoneme categories. Accuracy for plosives and affricates in children's speech improved more than 40% over baseline. Performance matched existing methods using approximately 13 min of labeled data, while approximately 45-60 min of labeled alignments yielded significant improvement.

Conclusion: The Wav2TextGrid tool allows alternate alignment workflows where the forced alignments, via training, are directly tailored to match clinical-grade, manually provided alignments.

Supplemental material: https://doi.org/10.23641/asha.28593971.

Purpose: During speech production, complex patterns of coordinated movements between sets of articulators (e.g., jaw and tongue, jaw and lips) form precise and consistent constrictions at distinct locations along the vocal tract, despite rampant contextual variation. Speech motor control research seeks to uncover basic principles of organization governing interarticulator coordination during constriction formation, assuming many degrees of freedom for controlling articulator movements. This study tested the hypothesis that the motor system reduces degrees of freedom and facilitates coordination by reliably controlling interarticulator timing.

Method: Ten talkers produced vowel-consonant-vowel (VCV) sequences, recorded using electromagnetic articulography, with variation in production rate and syllable stress. V was /ɑ/-/ɛ/, and C was alveolar /t/-/d/ or bilabial /p/-/b/. Timing relations between peak velocities of condition-specific sets of articulators were determined during oral closure and release for C: jaw, upper lip, and lower lip for bilabial constrictions and jaw and tongue tip for alveolar constrictions.

Results: During oral closing, the timing of articulator peak velocities was tightly coupled across scalar changes in rate and stress, such that timing variation in one articulator was accompanied by proportional changes in the timing of another articulator. In contrast, the timing of peak velocities was less tightly coupled during subsequent oral opening. The timing of peak velocities was also more reliably differentiated by rate/stress condition during oral closing than opening, indicating that speech articulator movements are temporally coordinated primarily based on the part of movement related to constriction formation, rather than its subsequent release.

Conclusion: Findings align with the view that stable interarticulator timing relations underlie the achievement of precise and consistent vocal tract constrictions during speech.

Supplemental material: https://doi.org/10.23641/asha.28791872.

Purpose: Despite evidence to suggest that rehabilitation can improve airway protection in people with Parkinson's disease (pwPD), rehabilitative therapies are underutilized. One newer treatment approach with growing evidence to support its efficacy is cough skill training (CST). To improve utilization of rehabilitations such as CST, it is important to understand patient treatment experience. Therefore, the aim of this study was to define treatment acceptability and satisfaction of CST in pwPD.

Method: Thirteen pwPD were consecutively recruited from a trial during which participants completed two in-person sessions of CST via spirometry over 2 weeks. A mixed-methods approach was used, whereby quantitative data (obtained from questionnaires) and qualitative data (obtained from semistructured interviews) were integrated to provide a holistic understanding of patient experience. Data collection and thematic analyses (of qualitative data) were guided by the Systems Engineering Initiative for Patient Safety framework.

Results: While treatment demonstrated adequate acceptability with a System Usability Scale median score of 70/100 (scores ≥ 70 indicate acceptability), integrative examination of patient experience revealed several themes that highlight barriers and facilitators to treatment acceptability-as they relate to CST tools (visualization, lip seal, and measurement accuracy) and tasks (practice targets, feedback, practice amount, and future training). Most (61.5%) participants were "satisfied" or "very satisfied" with CST. Four themes emerged related to satisfaction: treatment relevance, patient awareness and control over their disease, skill acquisition, and skill transference to real-life choking events.

Conclusions: In this study, patient responses indicated that small, personalized adaptations to equipment, training targets, and feedback may enhance their experiences, while education and personalized goal setting may enhance treatment relevance and perceived benefit. Patient perspectives and needs can inform the refinement of person-centered clinical implementation of CST and improve treatment uptake.

Supplemental material: https://doi.org/10.23641/asha.28887167.

Purpose: Voice disorders associated with vocal hyperfunction are some of the most common vocal pathologies. Certain personality traits are thought to be a risk factor for developing these disorders. The trait theory of voice disorders (TTVD) provided a unified framework to understand these relationships. This study re-examines the TTVD by adopting current theories and methods from personality science.

Method: This cross-sectional study recruited individuals diagnosed with primary muscle tension dysphonia (MTD) or a diagnosis associated with phonotraumatic vocal hyperfunction (PVH), that is, benign bilateral lesions of the lamina propria and bilateral or unilateral vocal fold polyp(s). Participants completed a contemporary personality battery. Data were analyzed via structural equation modeling and compared to vocally healthy controls.

Results: Several significant differences existed between vocally healthy controls (n = 416) and participants with MTD (n = 71) or PVH (n = 38). Compared to the controls, participants with MTD reported lower levels of Stability, Plasticity, Conscientiousness, Openness/Intellect, and the corresponding aspects of Industriousness and Intellect, respectively. Conversely, the MTD cohort was significantly higher in Neuroticism and its two corresponding aspects, Withdrawal and Volatility. Likewise, when compared to controls, the PVH reported significantly lower levels of Stability, Agreeableness, its aspect Politeness, and the aspect of Industriousness. Finally, compared to the MTD cohort, individuals with PVH were higher on Extraversion, specifically the aspect of Assertiveness, and lower on the aspect of Politeness.

Conclusions: The current study largely replicates the initial TTVD studies and updates them with a modern theory of personality. These results provide a strong foundation for future investigations to continue to study the relationships between personality traits and voice disorders.

Supplemental material: https://doi.org/10.23641/asha.28843052.

Purpose: Listening difficulty (LiD), often classified as auditory processing disorder (APD), has been studied in both research and clinic settings. The aim of this study was to examine the predictive relation between these two settings. In our SICLiD (Sensitive Indicators of Childhood Listening Difficulties) research study, children with normal audiometry, but caregiver-reported LiD, performed poorly on both listening and cognitive tests. Here, we examined results of clinical assessments and interventions for these children in relation to research performance.

Method: Study setting was a tertiary pediatric hospital. Electronic medical records were reviewed for 64 children aged 6-13 years recruited into a SICLiD LiD group based on a caregiver report (Evaluation of Children's Listening and Processing Skill [ECLiPS]). The review focused on clinical assessments and interventions provided by audiology, occupational therapy, psychology (developmental and behavioral pediatrics), and speech-language pathology services, prior to study participation. Descriptive statistics on clinical encounters, identified conditions, and interventions were compared with quantitative, standardized performance on research tests. z scores were compared for participants with and without each clinical condition using univariate and logistic prediction analyses.

Results: Overall, 24 clinical categories related to LiD, including APD, were identified. Common conditions were Attention (32%), Language (28%), Hearing (18%), Anxiety (16%), and Autism Spectrum Disorder (6%). Performance on research tests varied significantly between providers, conditions, and interventions. Quantitative research data combined with caregiver reports provided reliable predictions of all clinical conditions except APD. Significant correlations in individual tests were scarce but included the SCAN Composite score, which predicted clinical language and attention difficulties, but not APD diagnoses.

Conclusions: The variety of disciplines, assessments, conditions, and interventions revealed here supports previous studies showing that LiD is a multifaceted problem of neurodevelopment. Comparisons between clinical- and research-based assessments suggest a path that prioritizes caregiver reports and selected psychometric tests for screening and diagnostic purposes.

Supplemental material: https://doi.org/10.23641/asha.28907780.

Purpose: Reduced ability to raise vocal pitch has been associated with risk of aspiration in some populations. However, the mechanisms driving this association are understudied. This secondary data analysis aims to add to our knowledge on the shared and distinct biomechanics of swallowing and maximum vocal pitch elevation.

Method: We used existing data from the study of Venkatraman et al. (2020) on 10 healthy younger (age range: 19-23, M = 21) and eight older (age range: 65-79; M = 73) adults who completed maximum pitch elevation and swallow tasks under videofluoroscopy. A MATLAB tracking tool and computational analysis of swallowing mechanics was used to analyze five elements of pharyngeal swallowing biomechanics (anterior and superior hyoid excursion, laryngeal elevation, pharyngeal shortening, tongue base retraction, head/neck extension). Canonical variate analysis (CVA) determined differences associated with task and age. Post hoc discriminant function analyses (DFAs) compared the events between tasks in each group.

Results: CVA revealed that 63.9% of variance was accounted for by task (D = 3.46, p < .0001) and 35.5% by age (D = 1.92, p < .0001). Across age, DFAs indicated similar anterior hyoid excursion and laryngeal elevation between tasks, but greater superior hyoid excursion during swallows, replicating earlier findings. We also found greater base of tongue retraction during swallows than maximum pitch and greater pharyngeal shortening during maximum pitch elevation compared to swallows across groups (D = 5.38, p < .0001).

Conclusions: Maximum pitch elevation and swallowing require similar anterior hyoid and laryngeal excursion. Added to the novel finding of greater pharyngeal shortening during pitch elevation, we indicate that pitch glides may be a mechanism for targeting pharyngeal dysphagia and warrant further investigation.

Purpose: Pitch perception is important for speech sound learning, and reading acquisition requires integration of speech sounds and written letters. Many individuals with dyslexia exhibit auditory perception deficits that may therefore contribute to their reading impairment given that complex pitch perception is crucial for categorizing speech sounds. Given rising interest in music training as a reading intervention, understanding associations between prior music experiences and pitch perception is important. This study explored the relationship between pitch perception skills and reading ability in young adults with and without dyslexia with various levels of musical experience.

Method: Young adults (18-35 years old) with (N = 43) and without (N = 105) dyslexia completed two pitch perception tasks, reading assessments, and a survey reporting formal music training and childhood home music environment (HME).

Results: Participants with dyslexia performed worse than typically developing peers on both pitch perception tasks. Single-word reading was related to pitch perception in the typically developing group only. Childhood HME positively correlated with mode categorization and simple pitch discrimination in both groups. Formal music training was associated with performance on both pitch perception tasks in the typically developing group, and simple pitch discrimination in the dyslexia group.

Conclusions: Pitch perception deficits may interfere with complex acoustic categorization and persist in some individuals with dyslexia despite prior music experiences. Future research should investigate the link between pitch perception and phonological awareness in dyslexia and assess whether music interventions targeting these skills improve reading.

Purpose: The aim of the study was to describe the expressive sign vocabularies of a group of children learning Australian Sign Language (Auslan).

Method: The spontaneous signs of 44 children aged 3.0-6.8 years enrolled in one early-years bilingual-bicultural educational program were documented using a new approach, the Handshape Analysis Recording Tool, across a 2-year period. The resultant corpus was analyzed to determine the frequency of word classes including nouns, verbs, and adjectives.

Results: There were 3,003 Auslan tokens and 806 different sign types. Nouns, adjectives, and verbs were highly represented in this exploratory study, comprising 54.1%, 21.0%, and 15.8% of the entire corpus, respectively. Preliminary analyses indicated differences in the composition of Auslan vocabularies when compared with existing spoken English and American Sign Language data.

Conclusions: This exploratory study identified that the types of Auslan word classes used by this heterogeneous group of young learners included a high proportion of nouns and adjectives. While comparisons with past data are stated with caution as the composition of the child sample group was not controlled, there is preliminary support for earlier exposure and focused teaching of Auslan to facilitate the development of more varied expressive sign vocabularies.