联合使用助听器和增强电子耳蜗的使用者的中文语音识别

Enhancing Chinese Speech Recognition for Cochlear Implant Users by Using Hearing Aid in the Contralateral Ear

目的 使用电子耳蜗与助听器联合的声学模拟考察在语音成形噪声中的中文语音识别。方法 使用1增1减的信噪比自适应变化过程,测试了6个具有正常听力、使用中文语言的被试者在语音成形噪声中的中文句子识别门限(speechrecognitionthresholds,SRT)。结果 中文语音识别门限SRT随模拟系统中助听器的低通滤波截止频率变化的曲线呈现与低通滤波的幅频响应相似的形状。结论 只要助听器恢复的残存低频声学听力覆盖了说话人的基频范围,这些低频信息就能显著地帮助电子耳蜗使用者在语音成形噪声中识别中文语音,无论电子耳蜗的整体输入频率范围与助听器恢复的频率范围是否交叠。

Objective The present study investigated Chinese speech recognition m speech-shaped noise using an acoustic simmatlon of a cochlear implant （CI） in conjunction with a.hearing aid （ HA）. Method Speech recognition thresholds （SRT） of Chinese sentence recognition in speech-shaped noise were measured for 6 normal hearing Chinese-speaking subjects using a l-up/l-down adaptive procedure. Result The data curve of the SRT versus the low-pass cutoff frequency in the HA simulation showed a shape similar to that of a low-pass filter. Conclusion As long as the residual acoustic hearing covers the pitch range of the speaker, low frequency information restored by HA can greatly assist CI users in Chineses peech recognition in speech-shaped noise, no matter their frequecy ranges overlap or not.