UNLABELLED: This study determined the feasibility of performing a multicenter trial using quantitative SPECT myocardial perfusion imaging in patients with acute myocardial infarction. The feasibility was assessed by a cardiac phantom. METHODS: Twenty-two gamma camera systems in 19 laboratories were evaluated. Each laboratory performed nine studies on the cardiac phantom and performed quality control tests of system uniformity, collimator quality and gantry alignment on their gamma camera system. Defects simulating "hypoperfused" myocardium of differing amounts were placed in the myocardium for eight of the nine studies. Measured defect size was compared to true defect size. RESULTS: A total of 198 studies from 22 systems were analyzed. Three studies were technically inadequate. For all 22 systems, the average correlation coefficient between true and measured defect size was 0.992 +/- 0.009, with a range from 1.00 to 0.97. Three systems were rejected due to slopes of the regression line outside the limits 1.00 +/- 0.10 and mean errors > 5% in estimating defect size. The remaining systems had a correlation coefficient of 0.995 + 0.008 with an average slope of 1.00 +/- 0.04 and an intercept of 0.11% +/- 1.57%. The mean error in estimating defect size was 2.08% +/- 0.69%. CONCLUSION: The small interlaboratory variation and the close correlation with true defect size observed in a cardiac phantom indicate the feasibility of quantitative myocardial SPECT as a useful tool in multicenter trials evaluating therapy in acute myocardial infarction. Preliminary objective testing is required, however, to identify systems with technical deficiencies.
UNLABELLED: This study determined the feasibility of performing a multicenter trial using quantitative SPECT myocardial perfusion imaging in patients with acute myocardial infarction. The feasibility was assessed by a cardiac phantom. METHODS: Twenty-two gamma camera systems in 19 laboratories were evaluated. Each laboratory performed nine studies on the cardiac phantom and performed quality control tests of system uniformity, collimator quality and gantry alignment on their gamma camera system. Defects simulating "hypoperfused" myocardium of differing amounts were placed in the myocardium for eight of the nine studies. Measured defect size was compared to true defect size. RESULTS: A total of 198 studies from 22 systems were analyzed. Three studies were technically inadequate. For all 22 systems, the average correlation coefficient between true and measured defect size was 0.992 +/- 0.009, with a range from 1.00 to 0.97. Three systems were rejected due to slopes of the regression line outside the limits 1.00 +/- 0.10 and mean errors > 5% in estimating defect size. The remaining systems had a correlation coefficient of 0.995 + 0.008 with an average slope of 1.00 +/- 0.04 and an intercept of 0.11% +/- 1.57%. The mean error in estimating defect size was 2.08% +/- 0.69%. CONCLUSION: The small interlaboratory variation and the close correlation with true defect size observed in a cardiac phantom indicate the feasibility of quantitative myocardial SPECT as a useful tool in multicenter trials evaluating therapy in acute myocardial infarction. Preliminary objective testing is required, however, to identify systems with technical deficiencies.
Authors: Adelaide M Arruda-Olson; Véronique L Roger; Allan S Jaffe; David O Hodge; Raymond J Gibbons; Todd D Miller Journal: JACC Cardiovasc Imaging Date: 2011-05