A kind of integrated serial algorithms for noise reduction and characteristics expanding in respiratory sound

Abstract

The computer-based lung respiratory sound analysis, which can provide more information about the condition of lung station, has achieved a great development in recent years. However, the external noise in respiratory sound signal is a large restriction to the further promotion of this technique. In this paper, a kind of serial integrated de-noising algorithms which consist of a FIR band-pass filter and a modified wavelet filter and an adaptive filter, is proposed to suppress the noise in respiratory sound signals. The design of this kind of filter and its practical application are studied. The practical application in de-noise of the lung sound shows that this filter has a good de-noising effect and a good performance in outstanding the acoustic characteristics.