Abstract
Left ventricular diastolic function can be evaluated noninvasively from gated blood-pool studies. Two of the more useful indices of diastolic function are peak filling rate and time-to-peak filling rate. We have compared two analytic methods (third order polynomial and Fourier series) for determining these parameters. We evaluated both of these methods on a set of 24 subjects and a set of simulated curves. Although there was close agreement between the results obtained with the two methods, the polynomial fit requires operator selection of boundary points which influence the measured parameters and also requires an independent determination of the time-to-end-systole. The Fourier series method does not require operator intervention and determines all relevant quantities automatically from the single fitting procedure. Thus, we conclude that the Fourier series fit is less subject to methodologic bias and is, therefore, the preferred method for the evaluation of left ventricular diastolic function.
