Background:
Gating methods developed for electrocardiographic-triggered radionuclide ventriculography are being used with nonimaging detectors. these methods have not been compared on the basis of their real-time performance or suitability for determination of load-independent indexes of left ventricular function. this work evaluated the relative merits of different gating methods for nonimaging radionuclude ventriculographic studies, with particular emphasis on their suitability for real-time measurements and the determination of pressure-volume loops.
Methods and results:
A computer model was used to investigate the relative accuracy of forward gating, backward gating, and phase-mode gating. the durations of simulated left ventricular time-activity curves were randomly varied. three acquisition parameters were considered: frame rate, acceptance window, and sample size. twenty-five studies were performed for each combination of acquisition parameters. hemodynamic and shape parameters from each study were compared with reference parameters derived directly from the random time-activity curves. backward gating produced the largest errors under all conditions. for both forward gating and phase-mode gating, ejection fraction was underestimated and time to end systole and normalized peak ejection rate were overestimated. for the hemodynamic parameters, forward gating was marginally superior to phase-mode gating. the mean difference in errors between forward and phase-mode gating was 1.47% (sd 2.78%). however, for root mean square shape error, forward gating was several times worse in every case and seven times worse than phase-mode gating on average.
Conclusions:
Both forward and phase-mode gating are suitable for real-time hemodynamic measurements by nonimaging techniques. the small statistical difference between the methods is not clinically significant. the true shape of the time-activity curve is maintained most accurately by phase-mode gating.