ESTRO 2024 - Abstract Book

S3419

Physics - Dose calculation algorithms

ESTRO 2024

Guus Spenkelink 1 , Sophie Huijskens 1 , Jaap Zindler 2 , Marc de Goede 1 , Jim van der Star 1 , Jaap van Egmond 1 , Anna Petoukhova 1 1 Haaglanden Medical Center, Department of Medical Physics, Leidschendam, Netherlands. 2 Haaglanden Medical Center, Department of Radiation Oncology, Leidschendam, Netherlands

Purpose/Objective:

Accurate beam modeling measurements and verifications are crucial for dose engines in treatment planning software. Our study aims to compare ArcCHECK (a cylindrical phantom with diode array for patient QA) measurements and calculations between two dose algorithms in RayStation: Monte Carlo (MC) and Collapsed Cone (CC). In addition to the two dose engines, all configurations of ArcCHECK verification are compared: with and without cavity insert, as well as a heterogeneous versus homogeneous virtual phantom for calculations. This study includes both 6 MV and 6 MV flattening filter-free photon (FFF) beams, relevant for complex VMAT plans such as head-and neck and stereotactic lung treatments with heterogeneities, as well as plans for multiple brain metastases in one isocenter, involving highly modulated small fields. We aim to investigate performance differences between different combinations of ArcCHECK configuration and dose engine, providing insight into MC and CC algorithm strengths and limitations. For both dose algorithms (MC Dose, version 2.0 and CC Dose, version 5.7), dose models for 6MV and 6MV FFF photon beams of an Elekta VersaHD accelerator with Agility MLC were commissioned in-house in RayStation 12A. Table 1 provides a complete overview of the treatment fields in this study. For each test, three results were obtained: a physical measurement performed on a LINAC, a TPS calculation using MC, and a TPS calculation using CC. The ArcCHECK was used both with and without cavity plug for all irradiated fields. Furthermore, two types of ArcCHECK CT images, provided by the manufacturer, were analysed: a heterogeneous megavoltage CT image representing the ArcCHECK with various densities, and an artificial homogeneous image featuring only air and PMMA densities (see images in Fig1). For 6 MV FFF, no heterogeneity correction factors for VersaHD were available for the ArcCHECK without plug. Therefore, no measurements were done in this setting. Results are compared based on the mean gamma at 3% / 3mm dose criteria – a lower value indicates better fit between measurements and simulation. Material/Methods:

Table 1

Rectangular fields

6

MV, 3x3,

6

MV

FFF:

2x2,

5x5,

10x10,

20x20,

1x3, 2x5, 4x10, 12x30 cm 2

Clinical VMAT plans

6 MV: 3 x head-and-neck, glioma, brain

6

MV

FFF:

2

x

vertebra,

3

x

stereotactic

lung,

3 x multiple brain metastases in one isocenter

Results:

Mean gamma values on the inhomogeneous images are generally higher for MC as compared to CC for rectangular fields (Fig1, right column, blue vs red), as well as compared to MC calculations on the homogeneous phantom (Fig1,