The systematic nonlinearity exhibited in a comparison of 2 commonly used electrocardiographic lead-systems, when the net QRS axis-angle from one system is plotted against the other, is quantitatively analyzed. A family of nonlinear correlation curves is implemented using the ratio of 2 correction factors, f_c/f_o, as a parameter. Such curves, replacing linear regression lines, provide a more realistic representation of the actual problem of this type.
Between the Frank and the 12-lead systems, the value of f_c/f_o is found to be 0.53 for the frontal plane from a sample of 1, 000 subjects. The combination of correction factors, f_c=1.155 and f_o=2.18, is a possibility among countless others to make up the same f_c/f_o of 0.53. Appreciable improvement in axis correlation is demonstrated after the data are corrected by such factors.