Adaptive heart-noise reduction of lung sounds recorded by a single microphone

When analyzing lung sounds, heart sounds are an unavoidable and sometimes severely disturbing source of noise. An attractive noise-reduction scheme which is able to preserve the entire spectrum of the desired signal is adaptive filtering. However, such a scheme usually requires a second ´noise-only´ reference-signal. A new noise-reduction method which needs only a single microphone signal is presented. This algorithm uses adaptive filtering and therefore still preserves the entire spectrum. Results from experiments with healthy subjects indicate that heart-noise reductions by 24 to 49% can be achieved using this method.