Advanced search
1 file | 4.19 MB Add to list

Acoustic measurement of overall voice quality in sustained vowels and continuous speech

Youri Maryn (UGent)
(2010)
Author
Promoter
(UGent) and (UGent)
Organization
Abstract
Measurement of dysphonia severity involves auditory-perceptual evaluations and acoustic analyses of sound waves. Meta-analysis of proportional associations between these two methods showed that many popular perturbation metrics and noise-to-harmonics and others ratios do not yield reasonable results. However, this meta-analysis demonstrated that the validity of specific autocorrelation- and cepstrum-based measures was much more convincing, and appointed ‘smoothed cepstral peak prominence’ as the most promising metric of dysphonia severity. Original research confirmed this inferiority of perturbation measures and superiority of cepstral indices in dysphonia measurement of laryngeal-vocal and tracheoesophageal voice samples. However, to be truly representative for daily voice use patterns, measurement of overall voice quality is ideally founded on the analysis of sustained vowels ánd continuous speech. A customized method for including both sample types and calculating the multivariate Acoustic Voice Quality Index (i.e., AVQI), was constructed for this purpose. Original study of the AVQI revealed acceptable results in terms of initial concurrent validity, diagnostic precision, internal and external cross-validity and responsiveness to change. It thus was concluded that the AVQI can track changes in dysphonia severity across the voice therapy process. There are many freely and commercially available computer programs and systems for acoustic metrics of dysphonia severity. We investigated agreements and differences between two commonly available programs (i.e., Praat and Multi-Dimensional Voice Program) and systems. The results indicated that clinicians better not compare frequency perturbation data across systems and programs and amplitude perturbation data across systems. Finally, acoustic information can also be utilized as a biofeedback modality during voice exercises. Based on a systematic literature review, it was cautiously concluded that acoustic biofeedback can be a valuable tool in the treatment of phonatory disorders. When applied with caution, acoustic algorithms (particularly cepstrum-based measures and AVQI) have merited a special role in assessment and/or treatment of dysphonia severity.

Downloads

  • VOLLEDIG DOCTORAAT.pdf
    • full text
    • |
    • open access
    • |
    • PDF
    • |
    • 4.19 MB

Citation

Please use this url to cite or link to this publication:

MLA
Maryn, Youri. Acoustic Measurement of Overall Voice Quality in Sustained Vowels and Continuous Speech. Ghent University. Faculty of Medicine and Health Sciences, 2010.
APA
Maryn, Y. (2010). Acoustic measurement of overall voice quality in sustained vowels and continuous speech. Ghent University. Faculty of Medicine and Health Sciences, Ghent, Belgium.
Chicago author-date
Maryn, Youri. 2010. “Acoustic Measurement of Overall Voice Quality in Sustained Vowels and Continuous Speech.” Ghent, Belgium: Ghent University. Faculty of Medicine and Health Sciences.
Chicago author-date (all authors)
Maryn, Youri. 2010. “Acoustic Measurement of Overall Voice Quality in Sustained Vowels and Continuous Speech.” Ghent, Belgium: Ghent University. Faculty of Medicine and Health Sciences.
Vancouver
1.
Maryn Y. Acoustic measurement of overall voice quality in sustained vowels and continuous speech. [Ghent, Belgium]: Ghent University. Faculty of Medicine and Health Sciences; 2010.
IEEE
[1]
Y. Maryn, “Acoustic measurement of overall voice quality in sustained vowels and continuous speech,” Ghent University. Faculty of Medicine and Health Sciences, Ghent, Belgium, 2010.
@phdthesis{888156,
  abstract     = {{Measurement of dysphonia severity involves auditory-perceptual evaluations and acoustic analyses of sound waves. Meta-analysis of proportional associations between these two methods showed that many popular perturbation metrics and noise-to-harmonics and others ratios do not yield reasonable results. However, this meta-analysis demonstrated that the validity of specific autocorrelation- and cepstrum-based measures was much more convincing, and appointed ‘smoothed cepstral peak prominence’ as the most promising metric of dysphonia severity. Original research confirmed this inferiority of perturbation measures and superiority of cepstral indices in dysphonia measurement of laryngeal-vocal and tracheoesophageal voice samples. However, to be truly representative for daily voice use patterns, measurement of overall voice quality is ideally founded on the analysis of sustained vowels ánd continuous speech. A customized method for including both sample types and calculating the multivariate Acoustic Voice Quality Index (i.e., AVQI), was constructed for this purpose. Original study of the AVQI revealed acceptable results in terms of initial concurrent validity, diagnostic precision, internal and external cross-validity and responsiveness to change. It thus was concluded that the AVQI can track changes in dysphonia severity across the voice therapy process.
There are many freely and commercially available computer programs and systems for acoustic metrics of dysphonia severity. We investigated agreements and differences between two commonly available programs (i.e., Praat and Multi-Dimensional Voice Program) and systems. The results indicated that clinicians better not compare frequency perturbation data across systems and programs and amplitude perturbation data across systems.
Finally, acoustic information can also be utilized as a biofeedback modality during voice exercises. Based on a systematic literature review, it was cautiously concluded that acoustic biofeedback can be a valuable tool in the treatment of phonatory disorders.
When applied with caution, acoustic algorithms (particularly cepstrum-based measures and AVQI) have merited a special role in assessment and/or treatment of dysphonia severity.}},
  author       = {{Maryn, Youri}},
  isbn         = {{9789090249803}},
  language     = {{eng}},
  pages        = {{212}},
  publisher    = {{Ghent University. Faculty of Medicine and Health Sciences}},
  school       = {{Ghent University}},
  title        = {{Acoustic measurement of overall voice quality in sustained vowels and continuous speech}},
  year         = {{2010}},
}