TY - JOUR T1 - <strong>Dynamic PET Imaging of Myocardial Glucose Consumption Using MR-based Cardiac/Respiratory Correction: Human studies</strong> JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 177 LP - 177 VL - 60 IS - supplement 1 AU - Tao Sun AU - Yoann Petibon AU - Paul Han AU - Chao Ma AU - Jinsong Ouyang AU - GEORGES EL FAKHRI Y1 - 2019/05/01 UR - http://jnm.snmjournals.org/content/60/supplement_1/177.abstract N2 - 177Purpose: Motion associated respiration and cardiac function is a well-known source of image quality degradation in cardiac Positron Emission Tomography (PET) studies. The continuous movement of the myocardium in PET scanning introduces artifacts that alter quantification of tracer concentration in cardiac tissues and deteriorate diagnostic quality of image. In addition to its recognized potential in clinical cardiovascular imaging a unique feature of hybrid PET/MR is its ability of providing a robust solution to the problem of motion in PET. The aim of this study was to assess the impact of MR-based cardiac/respiratory motion correction in dynamic cardiac PET. Methods: The motion correction method comprises of four key components: (i) Surrogate signals for the cardiac and respiratory cycles that are acquired continuously during the dynamic PET scan and used to assign a cardiorespiratory phase index to each measured coincidence event. (ii) A deforming MR-based cardiorespiratory tissue motion model that assigns a given cardiorespiratory motion phase index to a specific 3-D motion vector field. The motion model is obtained by means of navigators, tagging and golden-angle-based radial MR acquisitions. (iii) A motion-dependent attenuation model which is generated by deforming the acquired attenuation map with the motion model. (iv) A motion-compensated PET reconstruction algorithm that incorporates the motion and attenuation models to produce a motion corrected image volume for each dynamic frame without loss of SNR. Human studies were performed to evaluate the proposed correction method. All scans were acquired on a whole-body PET-MR system (Siemens Biograph mMR). Three volunteers were recruited for the study. FDG (average radioactivity: 10 mCi) was administered intravenously and a 50-min acquisition in list mode was initiated. Dynamic PET and motion tracking MRI data were simultaneously acquired during free-breathing. Four venous blood samples were drawn at 30, 35, 40 and 45 min after tracer administration for deriving the plasma input function. The PET data for each dynamic frame were reconstructed in two ways: with the proposed MR-based motion correction (MC) method and without motion correction (NMC). Improvement in cardiac PET image quality was assessed by quantifying the thickness of the myocardium wall as well as the myocardium contrast-to-noise ratio in a late dynamic frame. The Patlak analysis method was applied to both MC and NMC image sets to calculate voxel-wise FDG uptake rate constants (Ki) in the myocardium. Results: Fig.1a shows that for subject 1, MC images depict overall higher activity concentration, and improved resolution compared to NMC images, as evidenced by sharper myocardium wall and improved delineation of papillary muscles. Fig.1b shows box plots that quantitatively analyze the effect of motion correction on activity estimation in different regions of the myocardium. For all three subjects, the impact of MC was most significant in septum regions, where myocardium contrast-to-noise ratio increased up to 22% and wall thickness decreased up to 18% compared to the NMC images. Fig.2a shows that for subject 1, MC yielded Ki images with higher resolution and higher Ki values in the myocardium. Fig2b shows box plots that quantitatively analyze the Ki for three subjects, which increased up to 17% after MC. Conclusions: MR-based cardiac and respiratory motion correction benefits static and dynamic FDG PET human scans. Analyses of static data indicate that motion correction produces images with higher myocardium activity concentration, lower wall thickness and improved resolution. For dynamic imaging, higher FDG consumption rates were found in the myocardium after MC for all frames. The method potentially can provide high quality images for many cardiac PET imaging applications. This work is under support of NIH funding R01HL118261. ER -