S2275
Physics - Intra-fraction motion management and real-time adaptive radiotherapy
ESTRO 2026
Conclusion: Intrafraction prostate motion is comparable for SABR and non-SABR fractionations. For 70% of SABR fractions the prostate motion is less than 3 mm. For 30% of treatment fractions prostate monitoring is required to keep the target within the treatment margin. For a minority of fractions prostate motion is greater than the target margin for the whole treatment time. References: 1.Bannai, M., Brown, A., Rumley, C., Squire, T. and Tan, A. (2025), Optimal Planning Target Volume Margins to Account for Intra-Fractional Prostate Motion Relative to Treatment Duration: A Study Using Real-Time Transperineal Ultrasound Guidance. J Med Imaging Radiat Oncol, 69: 395-401.2.A.K. Richardson, P. Jacobs (2017) Intrafraction monitoring of prostate motion during radiotherapy using the Clarity® Autoscan Transperineal Ultrasound (TPUS) system. Radiography. Volume 23, Issue 4. Keywords: Prostate, intrafraction motion, SABR Uncertainty of breathing motion surrogate tracking during online adaptive MR-guided radiotherapy of liver metastases Marcel Nachbar 1 , Jannik Kersting 1 , Julius Grunwald 1 , Monica Lo Russo 2 , Moritz Schneider 1 , David Mönnich 2 , Simon Boeke 2 , Cihan Gani 2 , Maximilian Niyazi 2 , Daniela Thorwarth 1 1 Section for Biomedical Physics, Department of Radiation Oncology, University of Tübingen, Tübingen, Germany. 2 Department of Radiation Oncology, University of Tübingen, Tübingen, Germany Purpose/Objective: With high soft tissue contrast and live cine imaging, stereotactic radiotherapy of liver metastases has become one of the most frequent treatments at the 1.5 T MR-Linac (Unity, Elekta), enabling gated treatments without implantation of markers. However, the high framerate needed for live cine MR-based imaging presents challenges, including reduced image contrast and artifacts. Consequently, the anatomical region used for target tracking during gating typically Mini-Oral 4715 deviates from the treated volume. For multiple metastases, only one target can be tracked. This tracking of surrogates may induce uncertainties due to the different respiratory motion of liver segments. Therefore, this study retrospectively evaluates the relative motion between liver segments in treated patients to incorporate this information into future margin recipes. Material/Methods: Forty patients treated at the MR-Linac were
retrospectively analyzed. For each of the eight liver segments (I–VIII), five patients were selected whose 2D cine images were centered on a target within the respective segment. For each patient, 200 frames (1.6- 2.5 images/second) were selected from the midpoint of a randomly chosen session. In the initial frame, all visible liver segments were annotated by a medical physicist in sagittal and coronal orientation (Figure 1). The annotated liver segments were propagated to the following frames using a self-developed modified optical flow. The calculated center of mass (COM) coordinate was utilized to calculate a time-specific relative offset between liver segments during the breathing cycle. For each segment combination, cumulative histograms of the relative offset were generated across patients. The reported offset was defined by the larger absolute value of the 5th and 95th percentiles, ensuring 90% coverage across all frames.
Results: The relative offset between different liver segments was derived 65 times in the coronal plane and 86 times in the sagittal plane. Based on the COM offset of the entire liver, the mean end-inhale to end-exhale displacement across all patients was 2.6 mm (LR), 5.2 mm (AP), and 11.2 mm (CC). The median COM offset between segments was 2.5 mm (LR), 2.4 mm (AP), and 3.7 mm (CC). For each segment combination, the heatmap in Figure 2 visualizes the corresponding derived values.
Made with FlippingBook - Share PDF online