Correction of catheter distortion effects on mean transit time--dye-dilution method.

Abstract

A method of reconstructing a dye-dilution curve which has been distorted by the use of a long catheter is presented. A digital computer approach has been used in the procedure because it allows the greatest flexibility for data computation and because digital computers are widely available at universities and medical centers. First, the transfer function (distortion function) is characterized in the Laplace transform domain and then transformed into a Z-transform format. Next, the input variable is written in terms of the output variable, and the resulting Z-transformed expression is translated into a delay operator form from which a Fortran program is written. Since noise is always present in the electrical output signal, it was found that filtering of raw data was required before a meaningful reconstruction of the dye-dilution curve could be attempted. A digital filter with a cutoff frequency of 1.5 cps was chosen. The reconstruction technique was then applied to a cardiac output curve. It was found that under the prevailing physical conditions the calculation of mean transit time without the removal of catheter distortion was in error by 42%. The calculation of blood volume shows, of course, a proportional error. The use of "high fidelity" dye-dilution curves in the diagnosis of cardiac disease is briefly discussed.