Detection and correction of patient motion in dynamic 15O-water PET MPI.

BACKGROUND: Patient motion constitutes a limitation to 15O-water cardiac PET imaging. We examined the ability of image readers to detect and correct patient motion using simulated motion data and clinical patient scans. METHODS: Simulated data consisting of 16 motions applied to 10 motion-free scans were motion corrected using two approaches, pre-analysis and post-analysis for motion identification. Both approaches employed a manual frame-by-frame correction method. In addition, a clinical cohort was analyzed for assessment of prevalence and effect of motion and motion correction. RESULTS: Motion correction was performed on 94% (pre-analysis) and 64% (post-analysis) of the scans. Large motion artifacts were corrected in 91% (pre-analysis) and 74% (post-analysis) of scans. Artifacts in MBF were reduced in 56% (pre-analysis) and 58% (post-analysis) of the scans. The prevalence of motion in the clinical patient cohort (n = 762) was 10%. Motion correction altered exam interpretation in only 10 (1.3%) clinical patient exams. CONCLUSION: Frame-by-frame motion correction after visual inspection is useful in reducing motion artifacts in cardiac 15O-water PET. Reviewing the initial results (parametric images and polar maps) as part of the motion correction process, reduced erroneous corrections in motion-free scans. In a large clinical cohort, the impact of motion correction was limited to few patients.

Hybrid SPECT Perfusion Imaging and Coronary CT Angiography: Long-term Prognostic Value for Cardiovascular Outcomes.

Purpose To determine the value of cardiac hybrid imaging, performed by combining SPECT myocardial perfusion imaging (MPI) with coronary CT angiography, as a long-term predictor for major adverse cardiac events (MACEs) (death, myocardial infarction [MI], unstable angina requiring hospitalization, coronary revascularization). Materials and Methods For this retrospective single-center study, 428 patients referred between May 2005 and December 2008 were classified according to hybrid imaging findings into the following groups: (a) those with stenosis of 50% or greater (at coronary CT angiography) with ischemia (at SPECT) in subtended territory (matched), (b) those with coronary CT angiography and/or SPECT findings in unrelated territories (unmatched), and (c) those with normal findings at coronary CT angiography and SPECT. End points were all-cause death or MI ("hard events") and a composite of MACEs. The Kaplan-Meier method was used to identify survival free of MACEs, and Cox proportional hazard regression analysis was used to determine independent predictors for MACE. Results During a median follow-up of 6.8 years, a total of 160 MACEs, including 45 deaths, were observed in the final study population (mean age, 62 years ± 11 [standard deviation]; 132 women). The annual hard event rate was more than fivefold higher for patients with matched findings (n = 46 [7.0%]) and was threefold higher for patients with unmatched findings (n = 113 [3.7%]) compared with that for patients with normal findings (1.2%; n = 216 [1.2%]; P < .001). The MACE rates were 21.8%, 9.0%, and 2.4% for matched, unmatched, and normal findings, respectively. A matched finding was an independent predictor for MACE and hard events. Conclusion In patients evaluated for coronary artery disease, cardiac hybrid imaging is an excellent long-term predictor of adverse cardiac events. A matched hybrid finding is associated with a high annual cardiac event rate. © RSNA, 2018.