Time-frequency domain filtering of evoked otoacoustic emissions
Time-frequency domain filtering of evoked otoacoustic emissions
复制标题
DOI：
10.1121/1.4751537
复制
发表时间：
2012-10-01
影响因子：
2.4
通讯作者：
Sisto, Renata
中科院分区：
文献类型：
--
作者：
Moleti, Arturo;Longo, Federica;Sisto, Renata
时域滤波是一种标准的分析技术，它用于分离失真产物耳声发射响应的两个主要矢量分量，利用它们不同的相频关系。在这项研究中，提出了一种基于连续小波变换的时频滤波技术，以克服时域滤波技术的固有局限性，并将其扩展到刺激频率和瞬态诱发耳声发射的分析。首先在理论基础上讨论了所提出的技术的优点，然后通过将其应用于合成和真实的耳声数据的分析来实际证明。结果表明，时频分析方法可以根据经验进行优化，从而有效地分离出耳声响应中与不同产生机制或不同产生位置相关的成分。聚焦于具有给定时间-频率特征的耳声响应的单个分量也可以显著改善信噪比，因为随机噪声贡献倾向于均匀分布在时间-频率平面上。(C)2012年美国声学学会。[http：//dx.doi.org/10.1121/1.4751537]
Time-domain filtering is a standard analysis technique, which is used to disentangle the two main vector components of the distortion product otoacoustic emission response, exploiting their different phase-frequency relation. In this study, a time-frequency filtering technique based on the continuous wavelet transform is proposed to overcome the intrinsic limitations of the time-domain filtering technique and to extend it also to the analysis of stimulus-frequency and transient-evoked otoacoustic emissions. The advantages of the proposed technique are first discussed on a theoretical basis, then practically demonstrated by applying it to the analysis of synthesized and real otoacoustic data. The results show that the time-frequency approach can be empirically optimized to get effective separation of the components of the otoacoustic response associated with either different generation mechanisms or different generation places. Focusing on a single component of the otoacoustic response with a given time-frequency signature may also improve significantly the signal-to-noise ratio, because the random noise contribution tends to be uniformly distributed on the time-frequency plane. (C) 2012 Acoustical Society of America. [http://dx.doi.org/10.1121/1.4751537]