TY - JOUR T1 - 3D reconstruction with an extended system model for the new Biograph mMR whole-body MR/PET system JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 1992 LP - 1992 VL - 52 IS - supplement 1 AU - Rebekka Kraus AU - Gaspar Delso AU - Xiaoyin Cheng AU - Sibylle Ziegler Y1 - 2011/05/01 UR - http://jnm.snmjournals.org/content/52/supplement_1/1992.abstract N2 - 1992 Objectives Spatial resolution in PET imaging can be improved by extended system modeling. This work extends an analytical model of the mMR’s geometry to include the following resolution-degrading effects: Tissue-dependent positron range in magnetic field (MF), non-collinearity, and inter-crystal scatter and penetration. The improved model is then incorporated and tested within a 3D-MLEM reconstruction framework. Methods The positron range distribution is based on Palmer and Brownell’s analytical model, integrating the annihilation density over the positron energy spectrum. The effect is incorporated in the system model as a 3D blurring kernel applied prior to forward-projection. As MR-based attenuation maps yield tissue type for each voxel, positron range can be selected accordingly. The other effects are combined in the system model as a radially-variant filter in sinogram space. The non-collinearity component is modeled as a Gaussian distribution dependent on the distance of the detectors in coincidence. The detector response including inter-crystal scatter and penetration effects is determined by Monte-Carlo simulation. Results The resulting system model was validated with GATE simulations using a model of the mMR including the 3T MF and the MR body coil. A set of point sources around the scanner axis was simulated for different tracer and tissue combinations. By including the MF in the simulation, the reconstructed images improve in spatial resolution (FWHM), e.g., from 4.2 to 4.0 mm (radial) and from 3.9 to 3.8 mm (tangential) for O15 in soft tissue. The improvement is more significant in lung tissue, from 5.7 to 4.2 mm and from 5.2 to 4.2 mm respectively. The positron range compensation now leads to FWHM improvements of at least 10% in soft tissue and 14% in lung. Applying the full system model further improves the FWHM by 5%. Conclusions The results prove that the extended system model is well suited to improve spatial resolution for the mMR. Work on the extension of the positron range model to the borders between different tissue types is ongoing ER -