Modified spectral subtraction for de-noising heart sounds: Interference Suppression via Spectral Comparison

Abstract

Heart disease remains the leading cause of fatalities in the western world. To address this many long-term systems have been developed including heart sound monitors or phonocardiograms (PCG). In daily life ambient acoustic noise interferes with the heart sound signal. This noise must be suppressed to allow effective monitoring of heart function. This paper proposes a new 2-channel signal processing technique, termed interference suppression via spectral comparison (ISSC) which is based on the 1-channel spectral subtraction algorithm, and aims at improving the quality of the recorded heart sound or PCG data. The proposed technique provides the best overall quality improvement when applied to six different heart sounds when compared to other noise cancellation algorithms namely: simple subtraction, adaptive least mean squares (LMS) and recursive least squares (RLS) filter and wavelet thresholding.