The purpose of the study was to evaluate the reliability of ejection fractions obtained from first-pass radionuclide ventriculography with a large field-of-view tomographic single-crystal gamma camera.
Methods: A SPECT camera had its electronics redesigned to improve counting efficiency and was equipped with an experimental ultra-high sensitivity collimator. Left ventricular ejection fraction (LVEF) was measured in 28 patients by 30 degrees RAO first-pass imaging and by "best septal view" LAO planar equilibrium radionuclide ventriculography on a conventional small field of view Anger camera. For 28 other patients, first-pass ejection fractions were compared to multicrystal gamma camera values. Visual analysis was performed to judge clinical acceptability of first-pass images for identification of wall-motion abnormalities.
Results: Linear regression analysis of first-pass against equilibrium ejection fraction demonstrated good correlation (r = 0.92; slope = 0.90; intercept = 3.8; s.e.e. = 6.4%). First-pass ejection fraction values also correlated linearly with multicrystal camera values for the left ventricle (r = 0.94; slope = 1.05; intercept = 1.3; s.e.e. = 5.3%). For a subgroup of 19 patients, single-crystal camera right ventricle ejection fraction demonstrated good correlation with multicrystal camera values (r = 0.82; slope = 1.15; intercept = 1.3; s.e.e. = 6.1%). Interobserver variability correlated as r = 0.99 for LVEF ejection fraction and r = 0.92 for RVEF. Chi-square analysis of single-crystal first-pass image visual scores versus those from the gated equilibrium acquisitions showed close agreement (p < 10(-8)).
Conclusions: The evaluated camera/collimator system measured left and right ventricular ejection fraction accurately. Lung frame correction and dual regions were superior to paraventricular background correction and a fixed end-diastolic region.