S2076
Physics - Image acquisition and processing
ESTRO 2026
Results: Registration proved crucial in mitigating motion and minimizing image coverage of the residual motion mask. For liver (skeleton), mean coverage reduced from 20% (21%) to 7%/4% (24%/16%) using registration setting S/D. This was consistent with Dice trends. The increased mask coverage for S observed in skeleton likely originates from the algorithm’s limited ability to preserve bone structures against sliding motion of adjacent soft-tissue organs. The subsequent application of bilateral filtering, connected-component analysis, and morphological operations on DIF yielded an optimized balance between mask size and artifact coverage while maintaining low mask fragmentation (Figure1B). Small dose deviations relative to the treatment plan calculated for reference conditions underlined the framework’s ability to address motion- induced artifacts, reducing maximal dose deviations for photons/protons from 2.9%/31% to 1.6%/1.8% and decreasing the volume receiving dose differences >0.5Gy by 82%/98% (Figure2A). Incorporating both motion states from cDECT into treatment planning stabilizes target coverage in the presence of motion, highlighting one potential clinical benefit of the framework (Figure2B).
Conclusion: The proposed method efficiently mitigated motion and preserved DECT information to generate electron- density images for both anatomical states. This can transform the technical limitation of cDECT into a potential advantage for motion-aware radiotherapy planning. Keywords: motion, motion-aware, dual-energy CT Proffered Paper 3705 Quantitative evaluation of a stack artifact correction algorithm for 4DCT imaging Mushawar Ahmad 1,2 , Jannis Dickmann 2 , Niklas Lackner 1,3 , Oliver J. Ott 1,3 , Rainer Fietkau 1,3 , Stefanie Corradini 1,3 , Christoph Bert 1,3 , Juliane Szkitsak 1,3 1 Departement of Radiation Oncology, Universitätsklinikum Erlangen, Erlangen, Germany. 2 Cancer Therapy Imaging, Siemens Healthineers AG, Forchheim, Germany. 3 Comprehensive Cancer Center Erlangen, EMN, Erlangen, Germany Purpose/Objective: Irregular breathing during 4DCT acquisition introduces motion artifacts that compromise both image quality and anatomical accuracy. A stack-artifact reduction algorithm called ZeeFree-RT was developed to address these issues. While initial evaluations suggest improved image quality, systematic assessment of
Made with FlippingBook - Share PDF online