Suppr超能文献

[人工耳蜗语音处理器中的噪声信号降低]

[Noise signal reduction in cochlear implant speech processors].

作者信息

Müller-Deile J

机构信息

Klinik für HNO-Heilkunde, Universität Kiel.

出版信息

HNO. 1995 Sep;43(9):545-51.

PMID:7591867
Abstract

Cochlear implant wearers suffer from an additional impairment of auditory-verbal communication due to environmental noise, similar to that experienced by many users of conventional hearing aids. To reduce this problem the mini-22 cochlear implant system has implemented a noise suppression function in the Mini Speech Processor (MSP) decreasing background noise-induced stimulation. The efficacy of this algorithm was investigated in 18 experienced postlingually-deafened adult cochlear implant users. By means of a computer-assisted speech intelligibility measurement unit, speech understanding was compared with and without the noise suppression function. Paired observations were conducted in noise as well as in silence. The recording of complete performance-intensity functions in terms of speech discrimination tested with the standardized Freiburg speech intelligibility test and a newly developed sentence test demonstrated the benefit of the noise reduction mechanism. Better understanding was achieved in silence in the normal speech processing mode of the MSP, whereas a significant improvement in noise resulted from the noise suppression. Considerable interindividual differences were found in the gain patients derived from using the noise suppression algorithm. The more often the S-position was used, the better the individual result with activated noise suppression. However, providing the cochlear implant patients with more information, as in the spectral peak ("SPEAK") stimulation strategy recently developed by Cochlear Corporation, could lead to higher speech discrimination scores in interfering noise. Paired comparison studies revealed that the new strategy added significantly to results of speech understanding. Average intelligibility rose 40% when using the new coding method instead of the MSP with activated noise suppression.(ABSTRACT TRUNCATED AT 250 WORDS)

摘要

与许多传统助听器使用者类似,人工耳蜗佩戴者会因环境噪声而在听觉 - 言语交流方面遭受额外的障碍。为减少这一问题,mini - 22人工耳蜗系统在迷你言语处理器(MSP)中实现了噪声抑制功能,以减少背景噪声引起的刺激。对18名有经验的语后聋成年人工耳蜗使用者研究了该算法的功效。借助计算机辅助言语可懂度测量单元,比较了有无噪声抑制功能时的言语理解情况。在噪声环境和安静环境中都进行了配对观察。通过标准化的弗莱堡言语可懂度测试和新开发的句子测试来记录完整的言语辨别性能 - 强度函数,结果证明了降噪机制的益处。在MSP的正常言语处理模式下,安静环境中能实现更好的理解,而噪声抑制则在噪声环境中带来显著改善。使用噪声抑制算法时,患者获得的增益存在相当大的个体差异。S位使用得越频繁,激活噪声抑制时个体的效果就越好。然而,像科利耳公司最近开发的频谱峰值(“SPEAK”)刺激策略那样,为人工耳蜗患者提供更多信息,可能会在干扰噪声中获得更高的言语辨别分数。配对比较研究表明,新策略显著提高了言语理解结果。使用新编码方法而非激活噪声抑制功能的MSP时,平均可懂度提高了40%。(摘要截断于250字)

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验