RESUMO
Background corrections applied on the left ventricular volume curve determined by the "gated blood pool"--method are based on an estimated rather than on a directly measured background. This imposes an uncertainty on the values determined from the volume curve, especially on the ejection fraction. A method which does not require background correction may be applied if all available measurement and evaluation facilities are utilized fully. High temporal and spatial resolution is of fundamental importance, permitting the exact determination of the time-dependent scintigraphic contour variations of the left ventricle during the mechanical action of the heart. A good criterion of the validity of the volume curves with respect to interfering background radiation is the ejection fraction calculated from these curves. The direct intercomparison of 10 ejection fractions obtained by an expanded "gated blood pool"-method, employing cardiac catheterization, immediately before a biplane laevocardiography demonstrated very good agreement. A small systematic underestimation of the ejection fraction by the nuclear method was observed. This understimulation shows that the influence of the true background is small if other interfering count rate contributions or methodical uncertainties are excluded systematically.