ESTRO 38 Abstract book

S941 ESTRO 38

EP-1746 GLAaS absolute dose calibration for iViewGT EPID with flat and FFF beams: multicenter experience E. Vanetti de Palma 1 , J.M. Perez 2 , S. Ren Kaiser 3 , G. Nicolini 1 1 Radiqa Developments, Medical Physics Team, Bellinzona, Switzerland ; 2 Hospital Universitario HM Puerta del Sur, Servicio de Radiofísica, Móstoles - Madrid, Spain ; 3 Fondazione Poliambulanza Istituto Ospedaliero, Servizio di Fisica Sanitaria, Brescia, Italy Purpose or Objective Intensity modulation radiotherapy requires the implementation of adequate pre-treatment QA programs to guarantee safe treatments. Amorphous silicon EPIDs are particularly suitable for this task, due to their main characteristics: high spatial resolution, large area, stability, dynamic range and real-time acquisition. The GLAaS algorithm showed to adapt to a variety of applications with a simple direct calibration process to convert raw detector readings into absorbed-dose-to- water; nevertheless, up to now GLAaS validations were limited to Varian EPIDs technology. To test its feasibility out of single manufacturer environment, in this study GLAaS algorithm was applied to Elekta iViewGT images, using both flat and FFF beams. Material and Methods Dosimetric data for 6MV, 10MV, 6FFF and 10FFF beams, delivered from 2 Elekta linacs both equipped with FFF (Synergy and Versa HD) belonging to 2 different institutes, were collected at maximum dose depth. Primary and transmitted radiation were correlated with iViewGT pixel reading and GLAaS algorithm was configured for both flattened and unflattened beams. For open and modulated fields, comparison respect to TPS dose maps (Monaco- MonteCarlo, RayStation-Collapsed Cone) was assessed through profiles and gamma analysis (local and global dose normalization criteria). Concerning modulated fields, 15 IMRT/VMAT plans were pooled from the clinical database of each institute, according to the following stratification: 5 breast, 5 H&N and 5 SBRT cases. Results As already observed for Varian EPIDs, the response of iViewGT was confirmed linear. Similarly, pixel-by-pixel response changes in time differentiating on field and/or segment sizes and beam quality basis, i.e. primary or transmitted through MLC radiation component, were properly modelled. FFF beams operated at high dose-rates showed saturation effect of the detector: in order to overcome such behaviour, attenuators were placed on the detector front cover during irradiation. The satisfactory GLAaS algorithm configuration for both the linac models allowed pre-treatment QA in a variety of cases, IMRT and VMAT, flat and FFF beams, small and large fields. Analysis performed in field area showed an average gamma agreement index > 95% (global 3%-3mm) and > 90% (local 3% - 3mm).

effectively reduces verification inaccuracy caused by backscatter from the affected Varian EPID arm. EP-1745 Performance of ArcCHECK based quality assurance in helical tomotherapy with TomoEdge technology B. Yang 1 , H. Geng 1 , W.W. Lam 1 , K.Y. Cheung 1 , S.K. Yu 1 1 Hong Kong Sanatorium & Hospital, Medical Physics and Research Department, Happy Valley, Hong Kong SAR China Purpose or Objective As a useful 3D quality assurance (QA) tool, ArcCHECK has gained wide application in helical tomotherapy (HT), but most of the studies were based on old models. This study is to report our experience in ArcCHECK based QA for HT with TomoEdge technique including the establishment of the action level. Material and Methods A total number of 303 clinical plans in different treatment regions in our hospital were retrospectively studied. The tomotherapy plan verification was conducted using the ArcCHECK diode array with an acrylic insert for placing an A1SL ionization chamber. DQA plans were created by situating the target at the center of ArcCHECK for point dose measurement and meanwhile, making sure the electronic part of the device was not under the main beam. Gamma analysis method was then used to quantitatively compare the dose measured by ArcCHECK and that calculated from the treatment planning system (TPS). The criteria of 3mm distance to agreement (DTA), 3% dose difference, 10% threshold and absolute dose comparison were chosen. According to AAPM TG 119, the recommended action level of gamma passing rate was calculated as AL= Gm -1.96σ, where Gm was the mean of the gamma passing rate and σ was the standard deviation (SD). Results Details of our data are shown in table 1. A value pitch of 0.287 was used for almost all cases except for some brain treatments with values of 0.143 and 0.215. Jaw width of 2.5cm and 5cm with dynamic technique took up 95.7% of total cases and only fixed setting was used for 1cm jaw width. Good point dose agreements between measurements and TPS data were obtained with a mean deviation of 0.75%. Absolute gamma comparison gave an averaged passing rate of 96.6% with the SD of 4.7%, which resulted in an action level of 87.4%. Cases with lower gamma passing rate appeared to be in thorax and abdomen regions, which were suspected to be related to high MF and off-axis induced diode over-response. Our limited data showed that the gamma passing rate could be larger than 92% after a new DQA plan was created by locating the ArcCHECK close to the isocenter.

Conclusion The clinical action level of treatment plan verification using absolute gamma comparison can be established at 87.4% with 95% confidence level. High MF values and large off-axis setup might be the possible reason for low absolute gamma passing rate in some cases.