摘要
为去除瞬态诱发耳声发射信号中的刺激伪迹 ,在不降低信噪比的同时保留瞬态诱发耳声发射信号中的有用信息。论文采用小波变换的方法 ,分解了人耳物理模型中通过线性累加所获取的数据 ,找出了在不同刺激强度下 ,刺激伪迹所在的尺度、时间段及去除刺激伪迹的方法 ,然后将此法用于去除受试耳中记录到的信号的刺激伪迹。与导出的非线性响应方法相比 ,此方法不但有效去除了记录信号中的刺激伪迹 ,而且在不降低信噪比的同时保留了瞬态诱发耳声发射信号的有用信息。
The stimulus artifact, which contaminates in the initial segments of transient evoked otoacoustic emissions and inhibits the examination of high frequency cochlear responses, is reduced using discrete wavelet transform methods. Model ear data is decomposed using linear synchronous averaging at different stimulus intensities, to determine the distribution of the stimulus artifact at different stimulus intensities, and to find methods to reduce the stimulus artifact. The cochlear response signal is decomposed and reduced using this method. Compared with the derived nonlinear response method, the proposed method not only reduces the stimulus artifact and enhances the signal to noise ratio in the response region, but also retains useful information from the evoked otoacoustic emissions in the high frequency range. This technique will improve clinical applications of otoacoustic emissions.
出处
《清华大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2001年第9期31-35,共5页
Journal of Tsinghua University(Science and Technology)
基金
国家自然科学基金资助项目 ( 39870 2 12 )
关键词
瞬态诱发耳声发射
刺激伪迹
小波变换
非线性响应
transient evoked otoacoustic emissions
stimulus artifact
wavelet transform
derived nonlinear response