ESTRO 36 Abstract Book

S789 ESTRO 36 _______________________________________________________________________________________________

considering all the techniques, only 4.4% of mean g mean tests resulted out of tolerance. In addition, removing class-2 errors, this percentage decreases to approximately 3%. Actually the workload of IVD procedures on 9 patients is 1 hour per day. Conclusion IVD performed using SOFTDISO assures: (i) a rapid response of dose delivery alert with a reduced workload; (ii) a large number of patients tested daily and (iii) for out- of- tolerance tests repeating IVD in the subsequent day, the possibility to verify the efficacy of the adopted corrections. EP-1477 Evaluating gamma-index quality assurance methods for Nasopharynx Volumetric Arc Therapy (VMAT) E.M. Pogson 1,2,3 , S. Arumugam 2 , S. Blake 1 , N. Roberts 4 , C. Hansen 5,6 , M. Currie 7 , M. Carolan 7 , P. Vial 2 , J. Juresic 2 , C. Ochoa 2 , J. Yakobi 2 , A. Haman 2 , A. Trtovac 2 , L. Holloway 1,2,3,4,8 , D.I. Thwaites 1 1 University of Sydney, Institute of Medical Physics- School of Physics- Faculty of Science, Sydney NSW, Australia 2 South Western Sydney Local Health District, Liverpool and Macarthur Cancer Therapy Centres, Liverpool, Australia 3 Ingham Institute, Medical Physics, Liverpool, Australia 4 University of Wollongong, Centre for Medical Radiation Physics- School of Physics, Wollongong, Australia 5 Odense University Hospital, Laboratory of Radiation Physics, Odense, Denmark 6 University of Southern Denmark, Faculty of Health Sciences- University of Southern Denmark- Denmark, Odense, Denmark 7 Illawarra and Shoalhaven Local Health District, Illawarra Cancer Care Centre, Wollongong, Australia 8 University of New South Wales, South Western Sydney Clinical School, Sydney, Australia Purpose or Objective Pre-treatment dose verification is often performed on dose measuring phantoms with some form of gamma evaluation. However it has been shown that the clinical relevance of a 3% and 3mm pass rate tolerance is questionable. The purpose of this study is to simulate machine errors of clinical significance for nasopharynx patients and test if these errors can be detected on a standard commercial phantom. In this study systematic errors including collimator rotation, gantry rotation, MLC shifts, and MLC field sizes are investigated. Material and Methods Ten retrospective VMAT patients were planned with a department protocol. Machine errors were deliberately introduced to all plans. Plans were modified by increments using Python to create simulated error plans; -5 to 5° for gantry and collimator angles and -5 to +5mm for MLC shift and MLC field size, considering each parameter separately. Simulated error plans (Dose error ) were compared to the original non-error plan (Dose Baseline ) utilising equation (1).

The smallest unacceptable error plan for each error type (including the Gantry (G), Collimator (C), MLC Shift (S), and MLC Field Size (F) error was delivered on an Elekta Linac and dose was measured using an ArcCheck. Gamma analysis was performed in SNCpatient version 6.6 utilising a global 3%/3 mm (10% threshold with correction off) gamma pass rate. Before measurement, the Linacs were tested for MLC, gantry and dose accuracy. Only one patient’s -5° gantry error was deemed unacceptable and subsequently measured, patient 2 with this error detected (gamma pass rate of 68.8%). Results The results for 10 patients are shown in Figure 1.

Figure 1. Gamma pass rate (%) for non-error (NE) plans and for deliberately introduced errors (including the Collimator (C), MLC shift (S), and MLC Field Size (F) error), where the latter are selected as exceeding dose tolerances by the smallest magnitude. *Patient 1 F error of -5 was not deliverable due to machine tolerances, hence an F error of -2 is utilised here. The global 3%/3mm gamma pass is able to detect the majority of unacceptable plans, however some MLC field size plans still pass. Decreasing the MLC field size by 1mm can result in significantly reduced dose to the PTV, which affects tumour control e.g. patient 3 MLC FS-1 passed, however this plan under-doses the PTV63Gy by -5.4% relative to the original non-error plan. Conclusion Not all deliberately introduced clinically significant errors were discovered for VMAT plans using a typical 3%/3mm (10% threshold with correction off) gamma pass rate. EP-1478 A split field beam model of Beam Modulator linear accelerator in Pinnacle treatment planning system M. Chandrasekaran 1 , S. Worrall 1 , M.K.H. Chan 1 , N. Khater 1 , C. Birch 1 1 University Hospital Southampton NHS Foundation Trust, Radiotherapy Physics, Southampton, United Kingdom

(1) All error plans doses were then recalculated in Pinnacle 3 . Plans were reviewed against acceptable tolerance limits. Plans were above tolerance and considered unacceptable if PTV D95%, Brainstem D1cc or spinal cord D1cc were beyond a ±5% deviation in dose. Additionally if either of the left or right parotid mean doses were beyond ±10%, this was also considered an unacceptable plan.