Lyapunov exponents for japanese vowels in normal adult and cleft palate speakers

Katsuaki Mishima, Asuka Sugii, Tomohiro Yamada, Tatsushi Matsumura, Norifumi Moritani

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)

Abstract

Objective: The purpose of this study was to explore the possibility of using Lyapunov exponents (LE) to evaluate the voice and speech in cleft palate (CP) patients. Subjects and Methods: Fifty healthy adults (25 males and 25 females) and 17 adult patients (10 males and 7 females) with repaired CP, who had no hypernasality, were enrolled. Five Japanese vowels were recorded at 44.1 kHz. Nasalance scores were measured simultaneously. The first LE was computed for each 1-second interval, and the mean for the first LE from all intervals (mLE1) was then calculated. The delay time and embedding dimension were estimated at 13, 40, 34, 27 and 22 points for the vowels /a/, /i/, /u/, /e/ and /o/, respectively, and at 8 points for all vowels. Results: The mLE 1s for all vowels did not differ between normal males and females. The mLE1s for /a/ were significantly higher both in males and females with CP than in normal individuals. The correlation coefficient between the mLE1s and nasalance scores for all vowels was not significant for both normal and CP subjects. Conclusion: The LE seemed to be related to pathological disorders of the vocal cords in CP patients, but was independent of resonance parameters.

Original languageEnglish
Pages (from-to)129-133
Number of pages5
JournalFolia Phoniatrica et Logopaedica
Volume63
Issue number3
DOIs
Publication statusPublished - Apr 2011

Keywords

  • Autocorrelation
  • Cleft palate speech
  • Lyapunov exponent
  • Nasalance score
  • Nonlinear dynamic analysis

ASJC Scopus subject areas

  • Language and Linguistics
  • Linguistics and Language
  • Speech and Hearing
  • LPN and LVN

Fingerprint Dive into the research topics of 'Lyapunov exponents for japanese vowels in normal adult and cleft palate speakers'. Together they form a unique fingerprint.

Cite this