S. Nekolla, M. Souvatzoglou, S. Schachoff, F. Graner, S. Fürst, A. Martinez-Möller, S. I. Ziegler, M. Schwaiger Cardiac Attenuation Correction for an Integrated MR/PET System: Initial Comparison to PET/CT Radiological Society of North America (RSNA 2011) Chicago, IL, USA, November 2011 (bib)
	Purpose Attenuation correction (AC) in integrated MR/PET is a new challenge as the MRI does not allow direct assessment of 511keV photon absorption. We evaluate a multi tissue segmentation approach based on Dixon MR sequences and compare quantitative cardiac uptake values between PET/CT and MR/PET acquired sequentially. Method and Materials: From patients undergoing oncological PET/CT, 15 subjects with significant myocardial uptake of 18F-FDG were subsequently examined using the first integrated MR/PET (Biograph mMR). MR attenuation maps were calculated from a two point breathhold DIXON MRI sequence. The data was used for tissue classification into air, lung, fat and soft tissue, which generated an AC map with added AC templates for the MR scanners bed and spine coils. CT AC was performed with a low dose CT. Iterative PET image reconstruction for both PET/CT and MR/PET was used. Cardiac tracer distribution was quantified using volumetric sampling after automatic alignment of PET/MR and PETCT to correct for patient motion and normalized to the area with the largest uptake. Total and regional uptake in the AHA segment model was tested for deviations from the line of identify using Passing-Bablok (PB) regression with 95% confidence interval (CI). Results: Temporal delay between PET/CT and MR/PET was 66±18 (min:27, max:127) minutes. Normalized FDG uptake for PETCT and PETMR was 65±13% and 67±16% (p=ns), respectively. Mean differences were 5.1±4.3% (global) and 3.5±8.9% (regional). PB regression for global values showed a slope of 1.03 [CI: 0.84; 1.45] and intercept of -4.2% [CI: -37%;4%.]. For all 255 segments, the PB regression calculated a slope of 1.06 [CI:1.00;1.15] and an intercept of -6.5% [CI: -12.4%;-2.5%]. Thus, no statistical deviations from the line of identity were found for global data. However we found modest regional differences between PET/CT and MR/PET. Conclusion: When using a three tissue segmentation for MR images to derive PET AC information for hybrid MR/PET of the thorax, very similar cardiac uptake values were calculated as compared to sequentially acquired data from a PET/CT. Clinical Relevance / Application: PET attenuation correction in the heart is the most complex situation for hybrid PET imaging in general and MR/PET in particular. We demonstrated its clinical feasibility for this new imaging device.
	This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each authors copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.