基于听觉场景分析的电子听诊器降噪算法研究

A Study of Ambient Noise Cancellation Algorithm of Electronic Stethoscope Using CASA

针对在医院和救护车等噪声环境下使用的电子听诊器,提出了一种与噪声的类型无关的、基于人耳听觉场景分析的心肺音降噪CASA算法。该算法利用双通道采集到的信号和噪声的能量的不同,模仿人耳提取双通道之间强度差(IID)特征,以此为依据来辨别信号和噪声。再利用声学掩蔽效应,对噪声进行掩蔽,实现目标心肺音的增强。与自适应NLMS滤波相比,CASA的效果要好16 dB以上。

Aiming at ambient noise cancellation for electronic stethoscope used in noisy hospital and ambulance, a two - mic sound enhancement algorithm based on CASA is proposed. The algorithm manages to segregate target heart sound from ambient noise using cues such as interaural intensity difference （IID） and auditory masking effect. Just like adaptive algo- rithm, it is useful to almost all kinds of noise. And experiment results show that CASA can increase signal noise ratio（SNR） by 10 dB more than NLMS.