George S. Osborne College of Audiology, Salus University, Elkins Park, Pennsylvania 19027, USA.
Ear Hear. 2010 Aug;31(4):505-14. doi: 10.1097/AUD.0b013e3181d99a52.
The overwhelming majority of test measures to assess cochlear implant (CI) candidacy, efficacy, and progress are based on speech perception. Nonlinguistic sounds, such as alerting and nonspeech human generated sounds, have received comparatively little attention, despite their central importance for daily living and environmental sound awareness. The purposes of this study were to develop and validate a beta test measure of nonlinguistic sound perception and to assess performance in CI users.
A beta test of nonlinguistic sound perception, referred to as the NonLinguistic Sounds Test (NLST) was developed. The NLST consists of 50 sound tokens distributed over five categories (animal, human nonspeech, mechanical/alerting, nature, and musical instruments). Both closed-set (category identification) and open-set (token identification) nonlinguistic sound perceptions were examined. Twenty-two postlingually deafened CI users (mean age, 59.4 +/- 10 yrs) were evaluated using common speech perception test measures (Hearing In Noise test and Consonant-Nucleus-Consonant words) and the NLST following a pilot study in which nonlinguistic sound tokens used were selected or eliminated.
The NLST was easily administered to all 22 CI subjects. An overall token identification score of 49 +/- 13.5% correct was obtained across all five categories. CI users were able to identify the correct category for 71 +/- 11.5% of tokens. A moderate correlation between speech perception and accuracy of nonlinguistic identification was found (r = 0.519, p = 0.016).
The results suggest that nonlinguistic sounds are difficult for CI users to perceive. The categorization and identification scores suggest that sounds with harmonic structure or sounds with repetitive temporal structure are easier for CI users to perceive. A further developed clinical version of the NLST may be a useful clinical test to measure CI performance and progress, and perception of nonlinguistic sounds should receive greater attention during postimplant auditory rehabilitation.
评估人工耳蜗(CI)候选者、功效和进展的绝大多数测试措施都基于语音感知。尽管非语言声音(如警报声和非言语人声)对于日常生活和环境声音感知至关重要,但它们得到的关注相对较少。本研究的目的是开发和验证一种非语言声音感知的测试方法,并评估 CI 用户的表现。
开发了一种非语言声音感知的 Beta 测试,称为非语言声音测试(NLST)。NLST 由 50 个声音令牌组成,分布在五个类别(动物、非言语人声、机械/警报、自然和乐器)中。测试包括封闭式(类别识别)和开放式(令牌识别)的非语言声音感知。22 名后天失聪的 CI 用户(平均年龄 59.4 +/- 10 岁)接受了常见的语音感知测试(噪声下听力测试和辅音-核-辅音词)和 NLST 测试,此前进行了一项试点研究,其中选择或淘汰了用于测试的非语言声音令牌。
NLST 易于对所有 22 名 CI 患者进行测试。在所有五个类别中,平均正确识别率为 49 +/- 13.5%。CI 用户能够正确识别 71 +/- 11.5%的令牌类别。发现语音感知与非语言识别的准确性之间存在中度相关性(r = 0.519,p = 0.016)。
结果表明,CI 用户很难感知非语言声音。分类和识别分数表明,具有谐波结构或具有重复时间结构的声音更容易被 CI 用户感知。进一步开发的 NLST 临床版本可能是一种有用的临床测试,可以衡量 CI 的性能和进展,并且在植入后听觉康复期间,应更加关注非语言声音的感知。