Quantification of left to right cardiac shunts by multiple deconvolution analysis.

A new method for quantification of left to right cardiac shunts was studied in 17 patients scheduled for cardiac catheterization who had also undergone radionuclide angiocardiography. The observed pulmonary transit curve was deconvoluted in two different ways: (1) by the superior vena caval ("bolus") time-activity curve, to yield the deconvoluted pulmonary transit curve, which represented the theoretical pulmonary transit curve with a perfect bolus injection, and (2) by the right ventricular time-activity curve, to yield the pulmonary transfer function, which represented the theoretical pulmonary transit curve with a perfect bolus injection an with no intracardiac shunts. The pulmonary transfer function was superimposed on the deconvoluted pulmonary transit curve, and the area A under it obtained. The pulmonary transfer function was then subtracted from the deconvoluted pulmonary transit curve. The pulmonary transfer function was scaled to fit the resulting shunt recirculation peak in the difference curve, and the area B under this scaled pulmonary transfer function obtained. Shunt size was quantified as the pulmonary (QP) to systemic (QS) flow ration QP/QS = A/(A-B). The method correlated closely with oximetry (r = 0.93). Use of this multiple deconvolution analysis technique provides accurate shunt quantification and reduces subjective operator decisions.