Document Type: Original Article
Department of Nuclear Medicine, Imam Reza Hospital, Mashhad University of Medical Sciences, Mashhad, Iran
Department of Medical Physics, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran
Introduction: Motion of the patient during myocardial perfusion SPECT could potentially results in false perfusion defects. The effect of different reconstruction methods on these artifacts is not studied. Clarification of the relation between the extent, severity and duration of motion with the resultant artifacts may be helpful in designing special soft wares for motion correction. This study is a preliminary evaluation of motion artifacts in different reconstruction methods. Materials and Methods: Normal myocardial perfusion SPECT from a patient with low risk of CAD and no motion during imaging according to the review of cinematic images by 3 nuclear medicine specialists is selected. The scan was acquired one hour after IV injection of 740MBq of Tc-99m-MIBI through 180 degrees from RAO 45 to LPO 45, in 32 frames and 30 seconds per frame. The original images were moved artificially in frames 7, 16 and 24 (as initial, mid and late portion of imaging) from 1 to 3 pixels and for 30-90 seconds using VISION software. Also this artificial motion was applied in X and Y axis and in negative and positive directions. One hundred forty four images were produced and all were reconstructed using filtered back projection (FBP) and iterative (OSEM) reconstructions and interpreted visually and semi-quantitatively (using 17 segment model with 5 point scoring). Results: Applying one, two and three pixels motion converted normal scan to abnormal image in 2.8%, 25% and 39.6% respectively using FBP and in 18.8% , 43.2% and 68.8% respectively using OSEM reconstruction.(P<0.001). Correlation of interpretation (Kappa) for two sets of images (FBP vs OSEM) was 0.488. Summed score (SS) for each scan reflects the intensity and severity of the defects. Overall, mean summed score was 5.25 in FBP and 8.08 in OSEM reconstruction (P<0.05). Mean SS in returning motions for 1, 2 and 3 frames(corresponding to 30, 60 and 90 seconds) were 2.52, 6 and 8.1 using FBP respectively and 3.85 , 8.77 and 11.58 in OSEM respectively (ANOVA, P<0.001). With non-returning motions no normal image was noted in both reconstruction methods. Conclusion: This study showed that OSEM reconstruction is more sensitive for motion artifacts compared to FBP. Also severity and extent of the defects were larger in OSEM reconstruction compared to FBP. Severity of the artifacts is increased with increasing duration of motion (or number of frames involved).