S1811
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
workflow efficiency without compromising dosimetric quality. These findings support the feasibility of delivering multi-target SRT using ring-based linacs in centres where C-arm systems are unavailable. Keywords: SRT, Halcyon, Coplanar Digital Poster 870 Multicenter assessment of CT calibration curves for Direct Density photon dose calculations Manuela Burghelea 1 , Vicki Trier Taasti 2 , Patrick Wohlfahrt 3 , Akos Gulyban 1 , Nick Reynaert 1 , Frederik Vanhoutte 4 , Evelien Bogaert 4 1 Department of Medical Physics, Université libre de Bruxelles (ULB), Hôpital Universitaire de Bruxelles (HUB), Institut Jules Bordet, Brussels, Belgium. 2 Danish Centre for Particle Therapy, Aarhus University Hospital, Aarhus, Denmark. 3 Cancer Therapy Imaging, Siemens Healthineers, Varian, Forchheim, Germany. 4 Department Radiation Oncology, Ghent University Hospital, Ghent, Belgium Purpose/Objective: The study investigates variations in CT number Look- Up-Tables (LUT) for the Direct DensityTM (DD) kernel across multiple institutions and evaluates the impact of harmonization based on a consensus guideline. Material/Methods: Twelve institutions, with sixteen Siemens Healthineers CT scanners, performed DD acquisitions of the Advanced Electron Density phantom (Sun Nuclear) following the consensus guide for LUT generation by Peters et al. [1]. The institute-specific clinical LUTs (LUTclin) were compared with the newly generated LUT (LUTguide). Differences in relative electron density (RED) and mass density (MD), calculated as percentages relative to water, were evaluated for four tissue categories: lung, adipose, soft tissue and bone. Results: Figure 1 shows the generated LUTguide for all CTs. Benefits of software version VB20, such as further reduced tube voltage (kV) dependency and enhanced beam hardening correction (BHC), were most evident for adipose tissue (high prevalence in-human) but also impacted the bone region. Generated VB20 LUTguide further substantiated DD’s concept by showing straight lines also for adipose tissue (Figure 1a). For the CTs with older software versions (i.e. all except goOpenPro1VB20 and 2VB20), harmonized RED LUTguide construction yielded average standard deviations (SD) of less than 0.005 for low-density tissues, while the average for bone SD was 0.01. This variation could be attributed to different X-ray tube spectra and BHC algorithms and slightly higher kV dependency in the earlier software versions (Figure 1b).
equivalent intra-target dose uniformity. GI values were consistent across methods with negligible bias relative to HyperArc. Increasing target number did not significantly degrade PCI or GI, while HI improved slightly with plan complexity, suggesting favourable modulation behavior. Correlation analyses indicated no clinically significant dependence of plan quality on PTV count, though weak non-linear trends were observed for modulation factor and GI in more complex plans.Fig1: Auto-v-Man differences and Compared to PTV Volume
Fig2: Bland Altman Auto V Manual NTO
Conclusion: When focused on target conformity parameters, coplanar SRT optimisation on ring-gantry systems can achieve clinically comparable conformity, homogeneity, and dose gradients to non-coplanar HyperArc plans for multiple brain metastases. Both manual and automatic NTO approaches were effective, with automatic optimisation offering
Made with FlippingBook - Share PDF online