Proceedings of The Physiological Society

University of Oxford (2011) Proc Physiol Soc 23, PC261

Poster Communications

Relationship Between Seismocardiography and Echocardiography for Measuring Cardiac Cycle Timing Events

J. Neary1, D. S. MacQuarrie1, V. Gebhardt2

1. Kinesiology & Health Studies, University of Regina, Regina, Saskatchewan, Canada. 2. Cardiology, Pasqua General Hospital, Regina, Saskatchewan, Canada.


Seismocardiography (SCG) has been studied for over 60 yr but has been impractical until recently because of advances in computer software and micro-processors. SCG records the low frequency vibrations induced by the heart, and records cardiac cycle mechanics and timing events. Echocardiography (Echo) is recognized as a standard for measuring cardiac performance but can be technically difficult and expensive. Limited data is available that has compared these techniques (1,2). Thus, the purpose of this study was to compare SCG with Echo to determine whether SCG could provide a valid measure of cardiac performance. We hypothesized that SCG would correlate with Echo cardiac timing events. Simultaneous SCG and Echo was performed on 28 (17 females) reportedly healthy participants (Mean ± SD; age=39.3±13.9 yrs; age range = 12-59 yrs). The SCG was recorded under resting conditions, with participants in the supine position, with a tri-axial accelerometer attached to the skin using solid gel electrodes placed over the sternum of the chest. Following SCG collection, standard Doppler and M-mode Echo measurements were collected with resulting images confirmed by an echocardiologist. Data analysis was performed on 10 cardiac cycle SCG’s and 1-3 Echo images per timing variable. The results showed the % differences between the Echo and SCG for all comparable variables ranged from 0.90% to 11%, with the coefficient of variation similar between methods. Independent T-tests (p<0.05) showed no statistically significant differences between the Echo and SCG, respectively, for aortic valve open to acceleration time (rapid ejection) of systole (71 ± 13 vs. 76 ± 12 msec), mitral valve open to E-wave (111 ± 33 vs. 110 ± 20 msec), E-wave to A-wave (370 ± 118 vs. 410 ± 144 msec), mitral valve open to E-wave (159 ± 20 vs. 164 ± 28 msec), and mitral valve open to mitral valve close (536 ± 129 vs. 573 ± 148 msec). Statistically significant differences were found for aortic valve open to aortic valve close (292 ± 18 vs. 264 ± 20 msec), and isovolumic relaxation time (81 ± 15 vs. 91 ± 8 msec). However, the Bland-Altman plot for aortic valve open demonstrated an absolute difference between the methods of 13.4 msec. These data suggest that SCG reliably measured cardiac performance when compared to Echo timing, suggesting that SCG could be used in applied and clinical research trials. Furthermore, the amplitude of the specific cardiac waveforms appeared to reflect the contractile nature of the heart.

Where applicable, experiments conform with Society ethical requirements