Correction for respiration artefacts in myocardial perfusion SPECT is more effective when reconstructions supporting collimator detector response compensation are applied.

PURPOSE: To assess the impact of respiration on myocardial perfusion imaging (MPI) SPECT processed with advanced algorithmic reconstructions.
METHODS: SPECT studies obtained from a phantom simulation and 49 respiratory-gated,one-day 99mTc-sestamibi scans were corrected for respiratory-related cardiac movement. Three types of reconstruction algorithms: (a) filtered back projection (FBP), (b) ordered subset expectation maximization in which collimator detector response was incorporated (OSEMCDR), and (c) OSEM-CDR with additional attenuation and scatter corrections (OSEM-CDRACSC) were applied to the corrected and uncorrected sets and analyzed quantitatively and qualitatively.
RESULTS: A discrepancy between the corrected and uncorrected bull's eye maps > or = 10% wasfound in 2%, 10%, and 20% of the FBP, OSEM-CDR, and OSEM-CDR-ACSC scans, respectively. In studies with more than 10-mm respiratory motion, the effect of motion was greater in OSEM-CDR and OSEM-CDR-ACSC datasets as compared to FBP processing.Qualitative and quantitative differences between corrected and uncorrected sets were significantly larger in OSEM-CDR and OSEM-CDR-ACSC data than in those of FBP data.
CONCLUSIONS: Respiratory-related cardiac motion significantly affects MPI-SPECT reconstructed with advanced high-resolution reconstruction algorithms such as OSEM-CDR and OSEM-CDR-ACSC and thus may justifies the application of respiratory gating.