RT Journal Article SR Electronic T1 Clinical evaluation of correction of cardiac respiratory motion accounting for irregular respiration JF Journal of Nuclear Medicine JO J Nucl Med FD Society of Nuclear Medicine SP 1364 OP 1364 VO 51 IS supplement 2 A1 Joyoni Dey A1 Michael King A1 P. Hendrik Pretorius A1 Ronn Walvick A1 Seth Dahlberg YR 2010 UL http://jnm.snmjournals.org/content/51/supplement_2/1364.abstract AB 1364 Objectives In SPECT when the respiration varies there can be a large variation in the counts in amplitude bins at different projection angles. This can result in limited-angle reconstruction artifacts which decrease the accuracy of motion estimation. The objective of this study was to evaluate the impact on clinical slices and polar-maps of a respiratory motion correction methodology we have developed which is formulated to overcome the effect of respiratory irregularity. Methods Our method employs only the projection angles with significant counts in common between the binned respiratory states when estimating the motion between them. To test the methodology cardiac SPECT studies were acquired in list-mode along with temporally synchronized signals from belts wrapped about the patients. The list-mode studies were binned as a function of the amplitude of the signal, selected angles for each binned state reconstructed, and the motion estimated. We then corrected for the motion in iterative reconstruction using all the counts for all the bins and the motion estimates across the bins. Evaluation consisted of comparing the corrected slices and polar-maps to those without correction for evidence of differences caused by motion correction. Results The number of counts per respiratory bin was seen to vary considerably with projection angle thus motivating the use of our algorithm. The extent of motion estimated was generally less than 1 cm, with superior / inferior motion the dominate direction for heart motion. The changes in the slices and polar maps were consistent with a decrease in smoothing as evidenced with a better separation of the heart from sub-diaphragmatic concentrations of activity, and clearer visualization of defects. These changes were best seen when iterative reconstruction included the modeling of system spatial resolution and no additional smoothing. Conclusions We have developed a respiratory motion correction methodology which diminishes the impact of respiratory motion in cardiac SPECT slices. Research Support NIH and Philips Medical Systems