| Abstract: | A semi-automated, variable-region-of-interest method of analysis was used to measure both global and segmental left ventricular (LV) and global right ventricular (RV) contraction with ECG-gated first-pass and equilibrium radionuclide ventriculography. Normal values were defined in 20 healthy volunteers, and in 24 symptomatic patients, the results were compared with right anterior oblique (RAO) contrast left ventriculography. The global LV ejection fraction (LVEF) obtained by equilibrium imaging in the left anterior oblique (LAO) projection correlated closely with the results obtained by the gated first-pass method in the RAO projection (r = 0.95) and those obtained with contrast left ventriculography (r = 0.94); furthermore, the interobserver variability was small (r = 0.985). The normal values for LVEF obtained using radionuclide techniques and contrast ventriculography did not differ, but with the equilibrium radionuclide method, the RV ejection fraction (RVEF) values were underestimated in comparison to those obtained by the RAO gated first-pass technique. In five patients with localised inferior segmental akinesis at contrast angiography, the RAO first-pass cine display demonstrated a corresponding wall-motion abnormality in all cases, but LAO equilibrium cine displays did so in only one out of five patients. For segmental quantitation of LV contraction, a computer programme defined the ventricular edge, divided the RAO LV images into five segments and determined both the segmental area contraction (SAC) and the counts-based segmental ejection fraction (SEF). Radionuclide SAC measurements correlated very strongly with SEF measurements (r = 0.94-0.99). Both radionuclide SAC and radionuclide SEF correlated well with contrast angiographic SAC, except in the inferobasal segment.(ABSTRACT TRUNCATED AT 250 WORDS) |
