Abstract
Two straightforward manual methods of calculating left ventricular ejection fraction were studied. Data were obtained from gated blood pool studies on 125 cardiac patients with myocardial infarcts of varying extent and location. Ejection fractions ranged from 10 to 76%. The results confirm that if simple rules are agreed and adopted for a small group of observers within a centre, then low inter-observer variability can be obtained over the full range of ejection fractions. The fixed region method provides a systematically low ejection fraction which may be acceptable in many circumstances. The moving region technique involves marginally more operator interventions but should produce ejection fraction estimates closer to the 'true value'. Both protocols enable left ventricular ejection fraction to be obtained quickly and consistently. The disparity between the two procedures is systematic but not constant and increases towards higher ejection fractions.