Dynamic Comparison of Current Ballistocardiographic Methods

Abstract

This paper will be presented in three parts, appearing in successive issues of Circulation. In Part I certain force artefacts arising from the method of supporting the body, are compared with the "true" cardiovascular forces. This is done for the methods of Burger, v. Wittern, Dock-Taubman, Smith and Walker. These are approximately single-mass systems in which platform effects prove negligible or absent; so that at the lower ballistocardiographic frequencies, the internal driving forces on the body are opposed by relatively simple constraints. At this stage also, we will examine the complementary meanings of displacement, velocity and acceleration records. The same record (e.g. "body displacement") can have various physical meanings, depending on what aspect of cardiovascular dynamics the reader wishes to refer to. If, however, resonance occurs in the spectrum (direct-body methods), it is shown why no clear physical meaning applies to any of the records: displacement, velocity, or acceleration. In Part II we show experimentally that a similar analysis in terms of two masses applies fairly well to the bed methods. This analysis shows how the platforms of Starr, Nickerson, Burger and v. Wittern affect the errors over a wider BCG frequency band. Principles and methods for minimizing these errors will be examined. In Part III we examine the validity of the standard assumptions of ballistocardiography, employed in Parts I and II. Using ballistocardiographic records taken with and without beds, and subtracting the effects of the supports under the usual assumption, we show that mechanical errors still remain. Constraints to ground as stiff as the direct-body and Starr supports seem to excite more errors of body complexity than do the softer supports of Nickerson and Burger or v. Wittern, beyond the obvious distortions from resonance. Conclusions are drawn about the direction of progress toward more informative and reliable BCG records.