Turgeon Christine, Trudeau-Fisette Pamela, Fitzpatrick Elizabeth, Ménard Lucie
Department of Linguistics, Université du Québec à Montréal, Montreal, QC, Canada; Center for Research on Brain, Language, and Music, Montreal, QC, Canada.
Department of Linguistics, Université du Québec à Montréal, Montreal, QC, Canada; Center for Research on Brain, Language, and Music, Montreal, QC, Canada.
Int J Pediatr Otorhinolaryngol. 2017 Oct;101:87-96. doi: 10.1016/j.ijporl.2017.07.022. Epub 2017 Jul 21.
In child cochlear implant (CI) users, early implantation generally results in highly intelligible speech. However, for some children developing a high level of speech intelligibility may be problematic. Studies of speech production in CI users have principally been based on perceptual judgment and acoustic measures. Articulatory measures, such as those collected using ultrasound provide the opportunity to more precisely evaluate what makes child CI users more intelligible. This study investigates speech production and intelligibility in children with CI using acoustic and articulatory measures. Ten children with unilateral or bilateral CIs and 13 children with normal hearing (NH) participated in the study. Participants repeated five English vowels (/a/, /e/, /i/, /o/, /u/) with and without auditory feedback. Ultrasound was used to capture tongue positions and acoustic signals were recorded simultaneously. The results showed that, despite quite similar acoustic results, the two speaker groups made different use of the tongue to implement vowel contrasts. Indeed, the tongue position was lower in the feedback OFF condition than the feedback ON condition for all participants, but the magnitude of this difference was larger for CI users than for their NH peers. This difference led to diminished intelligibility scores for CI users. This study shows the limitation of acoustic measurements alone and demonstrates how the use of articulatory measurements can explain intelligibility patterns. Moreover, our results show that when cochlear implantation occurs early in life and auditory feedback is available, CI users' intelligibility is comparable to that of their NH peers.
在儿童人工耳蜗（CI）使用者中，早期植入通常会带来高清晰度的语音。然而，对于一些儿童来说，达到高水平的语音清晰度可能存在问题。对CI使用者语音产生的研究主要基于感知判断和声学测量。诸如使用超声收集的发音测量方法，为更精确地评估是什么使儿童CI使用者的语音更清晰提供了机会。本研究使用声学和发音测量方法调查了CI儿童的语音产生和清晰度。10名单侧或双侧CI儿童和13名听力正常（NH）的儿童参与了该研究。参与者在有和没有听觉反馈的情况下重复五个英语元音（/a/、/e/、/i/、/o/、/u/）。使用超声捕捉舌头位置，并同时记录声学信号。结果表明，尽管声学结果相当相似，但两组说话者在利用舌头实现元音对比方面存在差异。事实上，在无反馈条件下，所有参与者的舌头位置都比有反馈条件下低，但CI使用者的这种差异幅度比听力正常的同龄人更大。这种差异导致CI使用者的清晰度得分降低。本研究显示了仅靠声学测量的局限性，并证明了发音测量的使用如何能够解释清晰度模式。此外，我们的结果表明，当在生命早期进行人工耳蜗植入且有听觉反馈时，CI使用者的清晰度与听力正常的同龄人相当。
