Umatani Masanori, Ogawa Makoto, Iwahashi Toshihiko, Hosokawa Kiyohito, Kato Chieri, Inohara Hidenori
Department of Otorhinolaryngology - Head and Neck Surgery, Osaka University Graduate School of Medicine, Suita, Japan.
Department of Otorhinolaryngology - Head and Neck Surgery, Osaka University Graduate School of Medicine, Suita, Japan; Department of Otorhinolaryngology, JCHO Osaka Hospital, Osaka, Japan.
J Voice. 2022 May;36(3):297-308. doi: 10.1016/j.jvoice.2020.06.002. Epub 2020 Jul 10.
This study aimed to assess the feasibility of a neck-surface accelerometer (ACC) for estimating the amount of acoustic output produced during phonation regardless of individual differences or the mouth configuration.
The sound pressure levels (SPL) and signals of a neck-surface ACC were simultaneously recorded, while each of healthy nondysphonic speakers was asked to perform the following phonatory tasks: (1) repetitive phonation of the vowel /e:/ at various loudness levels; (2) gradually increasing vocal loudness (crescendo) during sustained phonation of the vowel /e:/; (3) repetitive smooth transition between phonation of the vowel /a:/ and /u:/ with the same vocal effort or between phonation of the vowel /e:/ and production of a hum /m:/. The skin acceleration levels (SAL) were calculated from the ACC signals.
Although the correlations between the SPL and SAL values were nearly linear in both repetitive-vowel-phonation and vowel-crescendo tasks, the crescendo task showed a higher correlation within individuals than the repetitive task, but with substantial individual differences. The correlation between the increments in the SPL and SAL was higher than that between the SPL and SAL. In the smooth-transition tasks, the SAL-but not the SPL-showed no significant differences between the vowels or hum.
These results show that the signal intensity of a neck-surface ACC reflects the amount of acoustic output during phonation irrespective of the mouth configuration, but shows nonnegligible individual differences. The use of the increment in the SAL is suggested to be suitable for comparing the amount of acoustic output.
本研究旨在评估颈部表面加速度计（ACC）用于估计发声过程中产生的声学输出量的可行性，而不考虑个体差异或口部形态。
同时记录颈部表面ACC的声压级（SPL）和信号，要求每位健康的无发声障碍的受试者执行以下发声任务：（1）在不同响度水平下重复发元音/e:/；（2）在持续发元音/e:/时逐渐增加发声响度（渐强）；（3）以相同的发声力度在元音/a:/和/u:/的发声之间或在元音/e:/的发声和发嗡嗡声/m:/之间进行重复的平滑过渡。根据ACC信号计算皮肤加速度水平（SAL）。
尽管在重复元音发声和元音渐强任务中，SPL和SAL值之间的相关性几乎呈线性，但渐强任务在个体内部显示出比重复任务更高的相关性，但个体差异较大。SPL增量与SAL之间的相关性高于SPL与SAL之间的相关性。在平滑过渡任务中，元音或嗡嗡声之间SAL有差异，但SPL无显著差异。
这些结果表明，颈部表面ACC的信号强度反映了发声过程中的声学输出量，与口部形态无关，但个体差异不可忽视。建议使用SAL的增量来比较声学输出量。
