Wavelet transform analysis of the seismocardiogram

Abstract

Seismocardiography is a new, non-invasive technique developed for recording and analyzing cardiac vibratory activity. Recent studies of seismocardiograms (SCGs) have demonstrated that they are a potential technique for detecting coronary artery disease. In this paper, wavelet analysis, based on multiresolution signal decomposition, has been used to investigate the effects of physiological conditions (isometric exercise and hyperventilation) on the SCG in different frequency sub-bands, with respect to the rest SCG. The power of the SCGs was calculated in seven frequency sub-bands. Significant differences between rest SCG, isometric SCG and hyperventilation SCG, were found in various sub-bands. The wavelet transform enabled distinctions to be made between SCGs for different physiological conditions, and showed it to be a promising analysis tool in the field of SCG analysis.