Comparison of one vs two CTs for attenuation correction of both rest and stress SPECT data for myocardial perfusion imaging with Tc-99m Tetrofosmin

Objectives To investigate whether radiation exposure could be reduced for patients undergoing myocardial perfusion imaging (MPI) by using only a single CT for attenuation correction (AC) vs the standard acquisition of a CT for both rest (R) and stress (S) with Tc-99m Tetrofosmin.

Methods Retrospectively, 20 patients with abnormal MPI results were chosen. Our 1 day scanning protocol for R/S MPI uses any 1 of 3 SPECT/CT scanners with a CT at both R and S for AC. Both R-CT & S-CT were used for AC of R & S MPI to produce 3 sum difference score (SDS) groups: group 1, the standard method of processing with R-CT AC for R-MPI & S-CT AC for S-MPI; group 2, using R-CT AC for both S-MPI & R-MPI; group 3, using S-CT AC for both S-MPI & R-MPI. All MPI were analysed by a single operator using 4D-MSPECT v4.2 to produce SDS groups. Linear regression was used to compare the differences in SDS between groups. Group 1 processing was repeated blindly in order to find intraoperator variability. Our institutional review board approved the project.

Results For intraoperator variability, linear regression yielded excellent agreement: y = 0.98x + 0.17, R²=0.79*. There were no significant differences between the SDS of groups 1 & 3 (y = 0.81x - 0.26, R²=0.73*) which represents how a single S-CT performs as the only source of AC. Conversely, there were significant differences in SDS between groups 1 & 2 (y = 0.74x + 0.08 R²= 0.57*) which represents how the R-CT performs as the single source of AC. *All linear regression analysis demonstrated an ANOVA significance (F) of <0.01.

Conclusions These results show that a single CT obtained during S-MPI may be used for attenuation correction of both R/S MPI without producing a significant difference in SDS; thereby, potentially reducing patient exposure to unnecessary ionizing radiation. Sole utilization of the R-CT for attenuation correction of both S and R MPI may lead to significant errors in SDS