S1929
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
[1] Keall, Paul J., et al. International Journal of Radiation Oncology* Biology* Physics (2025).[2] Li, Muheng, et al. Physics and Imaging in Radiation Oncology (2025): 100806.[3] Xiao, Fan, et al. Medical Physics 52.11 (2025): e70106.[4] Thummerer, Adrian, et al. Medical physics 52.7 (2025): e17981.[5] Siebert, Hanna, et al. IEEE Transactions on Medical Imaging (2024). Keywords: Dose calculation, Data challenge, MRI
a multiplicity of training data from a given patient dataset. Participants will address dose calculation on CT and MRI for photons and protons. Performance will be evaluated at both the photon segment/proton pencil beam and full plan (weighted sum of segment/pencil beam predictions) by metrics such as the mean absolute error (MAE), dose volume histogram indices, gamma analysis and model 75 patient from the SynthRAD2025 datasets were identified with good pairing as assessed by visual inspection and will be used to generate publicly available ground truth doses from photon MLC segments and proton pencil beam dose calculations (planned at least 500 beams per patient and modality [3], up to 2000 [2] depending on computing resources). 40 patient datasets will similarly be used to generate a private test set. Data pairing will provide both CT- and MRI-based dose distributions for training and evaluation. Figure 1 illustrates proof of principle data generation. The challenge is expected to open in March 2026 and have results presented at MICCAI2026 in October. runtime. Results:
Digital Poster 3290 Comparision Monaco-Eclipse TPSs
Patricia Lopez-Blanco Diaz, Elena Gonzalez Gonzalez, Hector Gomez Perez, Beatriz Garcia Costa, Alejandro Garcia Romero, Alejandro Barranco Lopez, Sheila Calvo Carrillo, Celia Castan Guerrero, Francisco Javier Jimenez Albericio, Pablo Ortega Pardina, Raquel Castro Moreno, Pedro Ruiz Manzano, Sonia Serrano Zabaleta, Ana Paula Millan Armengol Physics and radiation protection, Hospital Clinico Universitario Lozano Blesa, Zaragoza, Spain Purpose/Objective: The treatment planning phase has an important impact on the quality of the radiotherapy process. There are different planning systems that use various optimization and calculation algorithms. At the Lozano Blesa University Clinical Hospital in Zaragoza, we have two of these systems: Monaco and Eclipse. This study compares different metrics between them for volumetric modulated arc therapy (VMAT) in prostate SBRT treatments. Material/Methods: The treatment planning system (TPS) Monaco v6.2 (Elekta) features a two-step optimization algorithm, first a pencil beam for fluence optimization and then dose calculation with XVMC (Monte Carlo),after configuring the segments with theMLC. Conversely,Eclipse v15.5 (Varian)employs a specific multiresolution algorithm for optimizationand calculates dose using AcurosXB.All analyzed plans use two full VMAT arcson theElekta VersaHD accelerator, modeled in both systems. The collimator settings were not modified in order to maintain consistency between both planning methods (Monaco combines the clockwise and counterclockwise arcs as a single beam). The prescription used was42 Gy in 6 fractions. Plans were optimized so that the relevant parameters of the resulting dose–volume histograms (DVHs) were similar in both systems, aiming to realistically meet the clinical objectives.Complexity indices such as the modulation complexity score (MCS)[1],plan modulation (PM)[2], and the monitor units per cGy per fraction ratio (MU/cGy) were evaluated, along with other relevant parameters (treatment duration, dose
Conclusion: DoseRAD2026 will provide a benchmark to determine the optimal approach for real-time dose calculation on both CT and MRI, allowing determining which approach is optimal in terms of speed and accuracy. References:
Made with FlippingBook - Share PDF online