Embedding a multichannel environmental noise cancellation algorithm into an electronic stethoscope

The paper describes a multichannel adaptive algorithm for the enhancement of cardiac sounds with respect to environmental noise. It combines sounds acquired from a couple of microphones to reconstruct the transfer function of a stethoscope head and its interaction with the patient’s body. This identification process allows to perform a distortion-less noise reduction. The filter is embedded into an electronic stethoscope, composed of a traditional acoustic head and an electronic section. This instrument allows to show on a display the heart sound and to store the acquisition into a removable media transferring data to a PC. A software tool able to reproduce, visualize, store and analyze cardiac sounds, for performing assisted diagnoses of cardiac diseases, completes the system. A demonstrator of the tool has been realized. Experimental results show significant improvements in noise reduction, when the filtering algorithm is applied.