|
KMID : 0390220190300010068
|
|
Journal of Clinical Otolaryngology, Head and Neck Surgery
2019 Volume.30 No. 1 p.68 ~ p.76
|
|
|
The Effect of Unvoiced Segments on the Cepstral Analysis in Age-Related Dysphonia
|
|
Kim Geun-Hyo
Lee Yeon-Woo
Bae In-Ho
Park Hee-June
Lee Chang-Yoon
Kwon Soon-Bok
|
|
|
Abstract
|
|
|
Background and Objectives：The purpose of this study is to investigate the effect of unvoiced segments on cepstral analysis in patients with presbyphonia.
Materials and Methods：A total of 381 participants (263 participants with presbyphonia and 118 with normal voices participated in this study. The sustained vowel /a/ 2 seconds and one sentence from ‘San-Chaek’ were edited and analyzed using the Praat script. Auditory-perceptual rating was also performed by three raters.
Results：All parameters except LHRatio_V and extracted continuous speech (EXT) showed a significant difference between the normal and the dysphonia groups. The normal group showed higher cepstral peak prominence (CPP) and smoothed CPP (CPPS) than the dysphonia group. Additionally, AP ratings of the normal group were lower than that of the dysphonia group. The correlation within the measured values ranged from -0.86 to -0.66. In continuous speech (CS), there was a significant difference of area under the curve (AUC) between CPP and CPPS, but there was no significant difference in EXT.
Conclusions： Through this study, we confirmed the effect of unvoiced segments on cepstral analysis. By removing the unvoiced segments, the measured values of the cepstral analysis showed an increase. Through further studies, we hope to increase the accuracy of discriminating dysphonia by extracting the unvoiced segments.
|
|
|
KEYWORD
|
|
|
Dysphonia, Acoustics, Aging, Praat
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|