A comparison of two quantitative approaches for correction of background counts of first-pass left ventricular time-activity curves: a numerical investigation

This report describes two quantitative approaches for correction of background counts for the determination of the left ventlicular ejection fraction from first-pass radionucide angiocardiography: one is based on the technique of complex demodulation and the other is based on the technique of spline interpolation. The least squares method is used as a paraneter estimation procedure. A piece-wise linear continuous function is adopted as the basic form of a back-ground curve estimate. The methods are investigated numerically using simulated examples in which the left ventricular time-activity curve is represented by mathematical functions. Numerical results show that the two methods have similar abilities to estimate the ejection fraction.