Design of a Heart Sound Extraction Algorithm for an Acoustic-Based Health Monitoring System

Abstract : The U.S. Army Research Laboratory is developing sensor technology to monitor the soldier's physiological variables and motor activities by gathering and analyzing acoustic data. The sensor/transducer consists of a fluid or gel contained within a small, conformable. rubber bladder or pad that also includes a hydrophone. This enables the collection of high signal-to-noise-ratio (SNR) cardiac. respiratory, voice, and other physiological data. The pad minimizes interference from ambient noise because it couples poorly with airborne noise. A heart sound interbeat interval extraction algorithm has been developed using only basic concepts and techniques generally known/available to first-semester Discrete-Time Signal Processing course graduates. The algorithm appears to yield good performance in environments where the in-band SNR is greater than roughly 10 dB and no more than one high-energy, in-band noise burst occurs within the timeframe of the IBI extraction "region of interest." The algorithm has been shown to be capable of effectively rejecting fairly intense breath sound events. Although no deliberate attempt was made to introduce voice signals into the acoustic environment, most of the energy content of the human voice (above approximately 80 Hz) would be eliminated by the 20- to 50-Hz linear-phase, band-pass filter utilized in this algorithm.