听觉诱发电位信号的小波消噪方法研究

Research on wavelet de-noising method of auditory brainstem response

本文对听觉诱发电位(auditory brainstem response,ABR)的信号消噪方法进行研究,采用Daubechies5(DB5)小波分析方法进行去噪处理。首先在MATLAB中对信号进行频谱分析,然后再利用小波工具箱找到DB5是适合于分析ABR信号的最优小波基,结合频谱分析结果提出小波系数的选取算法。研究表明选用DB5小波对测量信号进行变换处理,该分析方法不仅大大减少了叠加次数,同时也得到了更好的去噪效果。最后,在程序中实现了上述小波消噪算法,并依照临床感兴趣的Ⅰ-Ⅴ五波的波间潜伏期设计出波形评估程序,从而进一步证实了上述理论的可靠性。

This paper reports a research on denoising method for Auditory Brainstem Response（ABR）,in which DB5 wavelet analysis method is used for denoising.Firstly,frequency spectrum analysis of the ABR signal is made in MATLAB.Secondly,using MATLAB wavelet toolbox,DB5 is found to be the most appropriate wavelet basis for ABR signal;and combining with the results of frequency spectrum analysis,the selection algorithm of wavelet coefficients is proposed.This study shows that choosing DB5 to do wavelet transform on the metrical signal can not only reduce the superposing times greatly but also obtain better denoising result.Finally,the above mentioned arithmetic was realized by programming,a waveform evaluation procedure was designed according to the I-V inter-wave latency of clinical interest,which further substantiates the trustiness of the above mentioned theory.