Quantification of PET and CT Data Misalignment Errors in Cardiac PET/CT:Clinical and Phantom Studies

PET/CT units with high temporal resolution (particularly with 64-slice CT capability) are increasingly used as in clinical diagnosis and prognosis of cardiovascular disease. Since the CT sub-system in the combined PET/CT unit is used to perform attenuation correction of acquired PET data, misalignments between patient positioning for both scans can cause artifacts in the myocardial PET images potentially resulting in false positive artifacts. The aim of this study is to evaluate the misalignment effect (induced by spurious or physiological patient motion in-between the two modalities) on regional and global uptake values in the myocardial region. In this study, we used both phantom (RSD thorax phantom) and clinical studies (two FDG and one NH₃ rest/stress). Manual shifts between the CT and PET images ranging from 0 to 20 mm in six different directions were applied. Thereafter, attenuation correction was applied to the emission data using the manually shifted CT images in order to model patient motion between PET and CT. The reconstructed PET images using shifted CT images for attenuation correction were compared with the PET images corrected with the hypothetically misalignment free original CT image. The criteria and figures of merit used included VOI and linear regression analysis. The analysis was performed using 500 VOIs located within the myocardial wall in each PET dataset. The VOIs were uniformly distributed across all myocardial wall regions to assess the overall influence of PET and CT misalignment. The absolute percentage relative difference increased in all simulated movements with increasing misalignments for both phantom and clinical studies (up to 30% in some regions for the 20 mm shift). In conclusion, increasing the misalignment between PET and CT studies resulted in increased changes in the tracer uptake value within the myocardium both on a regional and global basis with respect to the reference as revealed by the various figures of meri- t used. The variation was more significant for right and down movements versus left and up directions.