S2208
Physics - Intra-fraction motion management and real-time adaptive radiotherapy
ESTRO 2026
Digital Poster 831 Characterization and analysis of Laplacian displacement vector fields for simulated tumor position shifts in lung cancer radiotherapy Nils Olovsson Department of Immunology, Genetics and Pathology, Uppsala University, Uppsala, Sweden. The Skandion Clinic, The Skandion Clinic, Uppsala, Sweden Purpose/Objective: Investigating and comparing the robustness of different treatment planning strategies and treatment modalities for lung tumors with regards to respiratory motion requires either extensive imaging or methods for simulating the motion. However, the imaging protocols that will allow for researching novel treatment methods might not be performed for patients as part of the clinical routine.Simulating an artificial motion with image deformations is an alternative when the available image data is limited. However, many methods for modeling image deformations are complicated by the need for generating and processing additional geometry and assessing material parameters that are not typically performed as part of the radiotherapy workflow.In a previous study we proposed a Laplacian displacement vector field (DVF) method for simulating small shifts in the position of lung tumors due to variations in breath-hold reproducibility (Olovsson et al 2025). The computation of the DVF can be performed directly on the image geometry and does not require any additional parameters. In this study the validity and limitations as well as the characteristics of such Laplacian DVFs were investigated. Material/Methods: An analytical solution to the Laplacian DVF was investigated in spherical geometry. It was compared to numerical solutions solved with finite differences on the same spherical test geometry as well as for geometry defined on a patient with lung cancer retrieved from an open dataset (Vandemeulebroucke et al 2007), Figure 1.The determinant of the Jacobian of the DVF transform was calculated for both the analytical and the numerical cases. A negative such value indicates a non-physical transform with spatial folding. The maximum tumor displacement allowed in order to have a physical DVF transform was derived in closed form for the spherical geometry and verified numerically on the patient geometry.
Conclusion: Higher breath-hold reproducibility uncertainty than anticipated during planning had only a small effect on the resulting target dose. Using 4D compared with 3D robust optimization did not result in better treatment plans. Nevertheless, substantial reduction in dose to organs-of-interest was demonstrated with proton therapy. References: Hoffmann, L. et al., 2023. Repeated deep-inspiration breath-hold CT scans at planning underestimate the actual motion between breath-holds at treatment for lung cancer and lymphoma patients. Radiotherapy and Oncology 188, 109887. https://doi.org/10.1016/j.radonc.2023.109887Josipovic, M. et al., 2019. Deep inspiration breath hold in locally advanced lung cancer radiotherapy: validation of intrafractional geometric uncertainties in the INHALE trial. BJR 92, 20190569. https://doi.org/10.1259/bjr.20190569Olovsson, N. et al., 2025. Robust treatment planning and displacement vector field editing for the probabilistic evaluation of proton radiotherapy of small lung tumors. Phys. Med. Biol. https://doi.org/10.1088/1361-6560/ae19c7 Keywords: SBRT, Proton therapy, Breath-hold
Made with FlippingBook - Share PDF online