Computerized analysis of equilibrium radionuclide ventriculography time-activity curve in the assessment of left ventricular performance: comparison of two methods.

Two methods for the analysis of left ventricle time-activity curve (TAC) of equilibrium gated ventriculography were compared in three groups of subjects [8 controls, 13 patients with coronary artery disease (CAD), 11 patients with myocardial infarction (MI). The first method was based on third-degree polynomial fitting, the second on Fourier analysis. The following parameters were calculated: peak ejection rate (PER), peak filling rate (PFR), time to PER and PFR, and filling fraction at the first third of diastole. A preliminary study of changing values of PER and PFR and of the mean error with increasing number of harmonics summed in order to obtain the best fitting of TAC demonstrated that beyond the sum of the first four harmonics there was no further significant improvement. The advantages of Fourier analysis are as follows: it is independent of the operator and fits only one function to the whole cardiac cycle; it requires less computer time; it provides better differentiation between controls and CAD patients. All of the 13 CAD patients had abnormal PFR on Fourier analysis, only 9 on polynomial analysis. At rest, 9 of the CAD patients had wall motion abnormalities, while only two had an abnormal ejection fraction.