S1920
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
values, indicating adequate fidelity for adaptive treatment planning. Nevertheless, lingering inaccuracies in low-density regions limit their utility for lung adaptive radiotherapy. Keywords: Enhanced CBCT, Adaptive planning Digital Poster 3186 Validation of Monte Carlo cell survival prediction in a preclinical setup Katiuska A. Pérez, Priscila Brandão, Jorge Jara, Sebastián Martínez, Andrea Russomando Física, Pontificia Universidad Católica de Chile, Santiago, Chile Purpose/Objective: This study aims to validate a computational framework capable of predicting cell survival from Monte Carlo simulations. The process integrates FLUKA [1], the Monte Carlo Damage Simulation (MCDS) code [2], and a mechanistic model [3] into a unified chain that links the physical process of energy deposition to the biological response, extending the methodology proposed in [4], from hadrons to photons. To this purpose, this work explores the applicability of the framework to a preclinical irradiation setup by reproducing the XRAD320 Precision system, widely adopted in preclinical radiobiology. Material/Methods: The irradiation geometry of the XRAD320 was reproduced in FLUKA v4-5.0 using manufacturer specifications and published data [5]. A comparison was performed with the experimental results reported by Claesson et al. (2011) [6], in which a 100 kVp X ray beam filtered with 1.7 mm Al was used to irradiate V79 cells. A comparable irradiation configuration, consistent with the experimental setup routinely used in our laboratory, was modeled to determine the secondary electron spectrum in the medium. The resulting spectrum was processed with MCDS to estimate DSB yields and complex lesion density ( λ ₚ ), which were subsequently used in a mechanistic model to derive SF(D). Bayesian inference was implemented in PyMC, sampling 100 posterior realizations. The results are reported as median SF values with a 16– 84% range.
density inserts ranging from 0.270 to 1.823 g/cm ³ was imaged using a Siemens Somatom Confidence CT simulator as the reference standard and an Elekta VersaHD CBCT system. Enhanced CBCT images were generated utilizing three commercial CBCT enhancement algorithms: Manteia ArtAssistant V0.1, RayStation version V2024A, and MIM version 7.0.3. HU accuracy was quantified via mean absolute error (MAE ± [SD] ） relative to the reference CT.Further validation of HU fidelity in the enhanced CBCT images was performed through dosimetric analysis in clinical cranial treatment plans for seven patients. In each case, the enhanced CBCT replaced the original planning CT, and doses were recalculated while maintaining identical isocenter coordinates and HU-to- electron density curves within the Monaco version 6.2.2 treatment planning system. Key dosimetric parameters, including the new conformity index (nCi), heterogeneity index (HI), and mean planning target volume (PTV) dose, were compared between the original plans and those derived from the enhanced CBCTs. Results: As detailed in Table 1, all enhanced CBCT modalities exhibited substantial HU accuracy improvements compared to the unprocessed CBCT for high-density inserts ( ≥ 1 g/cm ³ ). For instance, at 1 g/cm ³ , MAE decreased from 308.0±25.5 in the original CBCT to 0.0±32.6 (RayStation), 9.1±31.2 (Manteia), and 13.3±19.1 (MIM). Similarly, at 1.823 g/cm ³ , MAE reduced from 967.7±58.2 to 11.7±93.7 (RayStation), 114.9±101.4 (Manteia), and 29.9±77.0 (MIM). However, no consistent improvements were observed for low- density inserts (0.27 and 0.46 g/cm ³ ).In the dosimetric analysis of seven cranial patient cases, mean nCi values for recalculated plans were 1.26±0.24(RayStation),1.25±0.24(MIM), and 1.28±0.23(Manteia), aligning closely with the original plan value of 1.25±0.24. Comparable congruence was observed in HI and mean PTV dose metrics, as presented in Table 2.
Conclusion: Three evaluated CBCT enhancement algorithms substantially improve HU accuracy for materials approximating 1 g/cm ³ density. In cranial applications, enhanced CBCT images yield treatment plans of equivalent quality to those based on reference CT, as demonstrated by similar nCi, HI, and mean PTV dose
Made with FlippingBook - Share PDF online