S1936
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
ELM on TrueBeam and Halcyon, following the methodology of Van Esch et al, improving agreement between calculated and measured dose distributions. Enhanced modelling introduced by ELM could alter the representation of low dose regions and steep gradients, inherent in multi-metastatic SRT (MMSRT) planning, within the treatment planning system (TPS). This study evaluated the impact of ELM on these regions for SRT plans on the Halcyon system. Material/Methods: Ten MMSRT patients (2–6 lesions) were replanned using two calculation models: the standard Acuros XBv16 (XB16) and v18 (ELM). For each, two optimisation strategies were compared, manual and automatic normal tissue optimisation (NTO), to assess robustness across planning methods. The low-dose valley, defined as [Brain- ∑ PTVs], represented normal brain between targets. Dose-volume metrics were extracted across V50% to V2% of prescription dose to evaluate low-dose spread. Dose fall-off was quantified using concentric 1mm rings extending 1–7mm from the PTV surface, recording mean, median, and maximum doses within each ring. Clinically relevant spread was further evaluated using V12Gy, V20Gy, and V24Gy metrics. Statistical analysis was done using appropriate parametric or non-parametric testing
Conclusion: ELM introduces measurable yet clinically small changes in dose distribution for MMSRT on Halcyon treatment units, observed in low-dose and dose gradients. Whilst the changes were minor, the improved consistency and accuracy of dose modelling using ELM may strengthen user confidence in target dose conformity and organ-of-interest sparing, offering greater trust in calculated dose guidance for stereotactic applications on the Halcyon. References: Van Esch, Ann et al. “Testing of an enhanced leaf model for improved dose calculation in a commercial treatment planning system.” Medical physics vol. 49,12 (2022): 7754-7765. doi:10.1002/mp.16019 Keywords: SRT, Halcyon, Enhanced Leaf Model Digital Poster 3439 Evaluation of PlanAI: an artificial intelligent treatment planning assistant Paul J Doolan 1 , Constantinos Zamboglou 2,3 , Konstantinos Ferentinos 2,3 , Julie Shade 4 , Michael Bowers 4 , Peter Hoban 4 1 Department of AI, German Oncology Center, Limassol, Cyprus. 2 Department of Radiation Oncology, German Oncology Center, Limassol, Cyprus. 3 Department of Medicine, European University Cyprus, Nicosia, Cyprus. 4 Department of AI, Sun Nuclear Corporation, Florida, USA Purpose/Objective: Artificial intelligence (AI) is increasingly integrated into radiotherapy workflows1,2, but its role in treatment
(p<0.05). Results:
Analysis showed a consistent increase low doses to normal brain with ELM when compared to XB16, being most pronounced at lower dose levels (V10%–V2%), where volumetric exposure was significantly higher under ELM. Both manual and automatic optimisations showed this trend (Figure 1). The largest observed difference corresponded to an increase of ~35 cm ³ at the 1% dose level, which is clinically negligible.Target dose gradient analysis revealed statistically significant minor reductions in near-PTV doses, with maximum dose showing consistent decrease (Mean decrease 0.824Gy, 0.664sd), and increases in minimum dose (Mean increase 0.637Gy, 0.730sd) with distance (Fig 2). Volumetric doses in gradient regions (V12Gy to V24Gy) were lower under ELM, suggesting marginal steepening of dose fall-off.
Made with FlippingBook - Share PDF online