Correlation between subjective and objective measures of correspondence between different systems of vectorcardiography

Abstract

The agreement between vectorcardiograms obtained by different lead systems may be judged by various subjective and objective methods.

To test the validity of these methods, the results were studied when they were applied to vectorcardiograms obtained by the lead systems of Frank, McFee and Burger. For each individual case the three vectorcardiograms were compared pair-wise from the loop-projections (P) as well as from spatial wire models (S), and the degree of similarity was evaluated subjectively and expressed by a rating between 0 and 10.

Objectively, the difference between the vectorcardiograms was given by the “relative average true distance” (T) between isophasic points on the loops, as well as by the extent of the displacement brought about by a linear transformation designed to transform the one system into the other (δ). Correlations between P, S, T, and δ were calculated and were all found to be significant. T appeared to be interchangeable with δ (r = 1.00). Of the other correlations, the best one was between S and T (or δ) (r = 0.74), although the difference with P and T was only small. The regression coefficient proved to be plausible. For purposes of comparison of systems the use of wire models has only slight advantage over loop-projections.