ESTRO 38 Abstract book

S1009 ESTRO 38

I. Miloichikova 1,2 , Y. Cherepennikov 1 , B. Gavrikov 3 , A. Krasnykh 4 , S. Stuchebrov 4 1 Tomsk Polytechnic University, School of Nuclear Science & Engineering, Tomsk, Russian Federation ; 2 Cancer Research Institute of Tomsk National Research Medical Center of the Russian Academy of Sciences, Radiotherapy Department, Tomsk, Russian Federation ; 3 Moscow municipal oncological hospital №62, 1st Radiological Department, Moscow, Russian Federation ; 4 Tomsk Polytechnic University, Research School of High- Energy Physics, Tomsk, Russian Federation Purpose or Objective For electron beam radiation therapy of irregularly shaped tumors adjacent to critical organs, it is necessary to form complex fields with due consideration of the patients’ anatomical features. Consequently, there is a need to apply special tools in addition to standard applicators and blocks that come with a clinical electron accelerator. For example, customized complex-shape metal collimators produced by melting or cutting. This study proposes an alternative approach where polymer objects produced by rapid prototyping serve as electron beam-forming elements. The purpose of this study is to assess the applicability of the proposed approach. For this purpose numerical simulation of the clinical electron beams interaction with tissue-equivalent media, including polymeric materials suitable for the products creation by fused deposition modeling (FDM) was carried out. Material and Methods A numerical model simulating the external electron beam of a clinical accelerator and parameters of the polymer materials under study was developed on the basis of the GEANT4 mathematical and physical modeling libraries. For verification of the electron beam model, we used clinical dosimetry data of the ONCOR Impression Plus medical linear accelerator obtained in the 3D Scanner water phantom. To construct a model of polymer materials, we manufactured a set of products by fused deposition modeling from ABS, PLA and HIPS plastics and determined their actual density. The depth dose distributions of clinical electron beams in polymer objects were studied experimentally on the ONCOR Impression Plus accelerator using GafChromic EBT3 dosimetry films. Then we compared the results obtained with the calculated data. Results The developed numerical model of a clinical electron beam with near-real radiation parameters makes it possible to evaluate such characteristics of the depth dose distribution curve in a water phantom as R 50 (cm) with an accuracy of 0.05 cm, R 90 (cm) with an accuracy of 0.1 cm, d max (cm) with an accuracy of 0.2 cm and R p (cm) with an accuracy of 0.3 cm. The calculated and experimental data are in good agreement, which makes the numerical model suitable for determining the minimum thickness of a polymer absorber for developing collimating devices in the range of electron beam energies from 6 to 18 MeV. It was shown that the polymer thickness required for the absorption of the electron beam with an energy of 6 MeV is 4 cm, for an energy of 12 MeV, it is 8 cm, and for 18 MeV, 11 cm. The difference in thickness depending on the type of plastic (ABS, PLA and HIPS) is ±0.5 cm. Conclusion The numerical model of a clinical electron beam developed in this study allows estimating the distribution of ionizing radiation in tissue-equivalent media, exploring new approaches to the generation of irregularly shaped therapeutic fields with complex geometry, and predicting the dimensions to produce forming objects by fused deposition modeling.

EP-1858 Inter-observer variability in rectal target delineation on MRI for MR image-guided radiotherapy I. White 1 , A. Hunt 1 , T. Bird 2 , S. Settatree 3 , H. Soliman 3 , S. Bhide 3 1 Royal Marsden NHS Foundation Trust and Institute of Cancer Research, Academic Radiotherapy, london, United Kingdom ; 2 University Hospitals Bristol NHS Foundation Trust, Oncology, Bristol, United Kingdom ; 3 Royal Marsden NHS Foundation Trust and Institute of Cancer Research, Oncology, London, United Kingdom Purpose or Objective Evaluate inter-observer variability (IOV) in mesorectum and pelvic lymph node clinical target volume (CTV) delineation on 2D and 3D MRI sequences optimised for MR image-guided radiotherapy. Material and Methods Radiotherapy treatment planning MRI were acquired in 3 rectal radiotherapy patients. MRI sequences, which had been optimised for MR image-guided radiotherapy on a 1.5T Siemens radiology scanner, included high resolution 2D T2W and 3D T2W sequences with field of view to encompass the entire pelvis. 5 experienced radiation oncologists in rectal radiotherapy contoured the mesorectum, pre-sacral and right and left pelvic lymph node (LN) CTV regions of interest (ROI) on each 2D and 3D MRI sequence using Elekta MR Linac consortium rectal contouring guidelines. A simultaneous truth and performance level estimation (STAPLE) (1) was created for each ROI, on each MRI data set using Monaco ADMIRE software v2.0 (Research version, Elekta AB). Observer contours on each image data set were compared to the STAPLE contour by calculating the Dice similarity coefficient (DSC) and the mean and maximum distance to agreement (DTA). Analysis between MRI sequences was performed using Wilcoxon signed-rank test with GraphPad Prism, p<0.05 was considered statistically significant. Results IOV for the different target volumes on 2D T2W MRI and 3D T2W MRI is illustrated in table 1. There was very good agreement between observers and the STAPLE reference contours for both MRI sequences (Figure 1). DSC >0.8 was observed for all CTV ROIs, except for the presacral LN region when delineated on the 2D T2W MRI sequence (DSC 0.69).