S1798
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
recalculation. Material/Methods:
Twenty prostate cancer patients treated with moderately hypofractionated radiotherapy (20 fractions) were retrospectively analyzed. The workflow (Figure 1) uses daily CBCT imaging and rigid registration combined with AI-based auto-contouring to map planned baseline dose (derived from the planning CT) onto daily anatomies, enabling fraction- wise and accumulated dose estimation for prostate, bladder, and rectum. Anatomical variations were quantified through organ volumes, prostate center-of- mass shifts, and Dice similarity coefficients (DSC). Workflow performance was validated against CBCT- based dose recalculations in representative cases spanning minimal to maximal anatomical variation.
Figure 2. Example of accumulated DVH assessment. Top: Daily (thin) and planned (black) DVHs for bladder (blue), rectum (green), prostate (red), and prostate+1 cm (orange). Bottom: Comparison of accumulated (solid) vs. planned (dashed) DVHs for all structures. Conclusion: This workflow offers a robust, time-efficient solution for monitoring interfractional dose variations in prostate radiotherapy. By avoiding DIR and dose recalculation, it enables clinically feasible integration into routine workflows and provides quantitative support for offline adaptive strategies and toxicity risk Huang T-C, et al. Fractionated changes in prostate cancer radiotherapy using cone-beam computed tomography. Med Dosim 2015. doi.org/10.1016/j.meddos.2014.12.003.[2] Fuchs F, et al. Interfraction variation and dosimetric changes during image-guided radiation therapy in prostate assessment. References: [1] cancer patients. Radiat Oncol J 2019. doi.org/10.3857/roj.2018.00514.[3] Liang X, et al. Automated contour propagation of the prostate from pCT to CBCT images via deep unsupervised learning. Med Phys 2021. doi.org/10.1002/mp.14755.[4] Ong A, et al. Application of an automated dose accumulation workflow in high-risk prostate cancer - validation and dose-volume analysis between planned and delivered dose. Med Dosim 2022. doi.org/10.1016/j.meddos.2021.09.004. Keywords: Dose accumulation,AI contouring,Prostate RT
Figure 1. Automated workflow for dose estimation. Schematic of the pipeline combining CBCT imaging, rigid registration, AI-based contouring, and DVH calculation. The workflow runs in ~1 minute per fraction on standard hardware. Results: The full process runs in ~1 minute per fraction on a standard desktop computer. Prostate volumes remained stable, whereas bladder showed the largest fluctuations (+232% to -89% vs. baseline) and rectum intermediate variability. The prostate center-of-mass shift averaged 3.5 ± 2.1 mm with mean DSC 0.83 ± 0.08. Bladder dose metrics (V50Gy, mean dose) exceeded a 2% deviation threshold in nearly half of fractions. Validation against full CBCT-based recalculations confirmed agreement within ±2% for key parameters, even in patients with pronounced anatomical changes or CBCT artifacts (e.g., hip prosthesis). Figure 2 illustrates a representative accumulated DVH comparison for one patient.
Made with FlippingBook - Share PDF online