Effect of acquisition and analysis routines on gated blood pool measurements

Results obtained from gated equilibrium blood pool (GBP) studies are not only dependent on intrinsic variations, but also on the way in which images are acquired and analysed. The aim of this study was to investigate factors which could affect left ventricular time-activity curves. Temporal resolution was studied by comparing studies of 20 and 40 frames beat-1. Forty frames per beat resulted in a mean left ventricular ejection fraction of 0.48 compared to 0.46 for 20 frames beat-1. The mean difference of 0.02 was significant (P less than 0.01) as was the mean difference in maximum emptying rate (MER = 0.28, P less than 0.01) and in maximum filling rate (MFR = 0.38, P less than 0.01). No significant differences in ejection fraction (EF) values were found between acquisitions made in list and frame mode, but the mean differences for MER = 0.03 (P less than 0.05) and MFR = 0.01 (P less than 0.02) were significant. For patient repositioning and intra-observer variations no significant differences were found. In patients with normal EF values (greater than 0.5) no significant differences were found in the inter-observer study. In patients with anterior myocardial infarction (AMI), significant differences were found in EF, MER and MFR (EF = 0.02, P less than 0.001; MER = 0.2, P less than 0.01; MFR = 0.24, P less than 0.01). Significant differences were found in all values when comparing a semi-automatic method of evaluation with two automatic methods. In conclusion the results from this study suggest that acceptable reproducibility can be achieved in GBP studies, provided the method of analysis is not changed between studies.