[The determination of the ejection fraction of the left ventricle by the first passage of an indicator bolus. Comparison of two curve-fitting algorithms (author's transl)]

Abstract

In 28 studies of the first left ventricular bolus passage, the end-systolic and end-diastolic phase were determined in the left ventricular time activity curve (4 msec per frame) by simply smoothing this curve 25 times. Comparing this method with a highly sophisticated curve-fitting algorithm (Sinusoidial Approximation), a very high correlation (r = 0.95) was found of the resulting ejection fractions. Thus, the two methods can be regarded as equivalent. In comparison with the ejection fractions evaluated by cineventriculoangiographic studies, a good correlation (r = 0.85) was obtained as well. To generate end-systolic and end-diastolic frames the two frames containing the highest left ventricular radioactivity were designated as end-diastolic, and the two frames containing the lowest radioactivity as end-systolic, for each heart cycle. This criterion was found to be very efficient although the temporal resolution was 8 msec per image. Furthermore, mean filling- and ejection-times of the left ventricle were derived from the extensively smoothed time activity curve. Those parameters seemed not to be of great clinical value. By employment of fast computing procedures, the evaluation time of a first passage (including image and curve generation) could be reduced to less than 10 min. Accuracy remained constant so that this method was introduced into our clinical routine diagnostics.