ESTRO 37 Abstract book

S980

ESTRO 37

results in terms of dose distribution and MU calculation. Finally, regarding the QA method, we can finish up by saying that the accuracy and speed of 3D dose calculations have made it the preferred quality control method for pre-treatment verification in our institution. EP-1818 TrueBeam HDMLC DLG: Relationship between measured value & modeled Eclipse value in dose measurements G. Beyer 1 , C. Stacey 2 , S. Rizkalla 3 , R. Bodey 2 , K. Blythe 3 , N. Hindocha 2 , T. Greener 3 , U. Johnson 2 1 Medical Physics Services Intl Ltd, Medical physics Services intl. Ltd, Cork, Ireland 2 University College London Hospitals, Radiotherapy, London, United Kingdom 3 Guy’s and St. Thomas’ Hospital, Radiotherapy, London, United Kingdom Purpose or Objective The purpose of this work is to evaluate the relationship between the Dosimetric Leaf Gap (DLG) measured for HDMLC on TrueBeam STx (TBSTx) and the DLG value used in Eclipse modeling. Recent data has shown that a larger DLG value may be needed in Eclipse to obtain a good agreement between measured and predicted dose for RapidArc (RA) plans that results in dose discrepancies for fluence patterns and complex or small field IMRT plans. This work describes the results of opening slightly the physical DLG for HDMLC so that a value close to the measured DLG can be used in Eclipse thus producing close agreement between measured and predicted dose for dynamic fluence patterns, IMRT, and RA plans. Material and Methods The measured value for TBSTx DLG has decreased since the original machine release due to software changes. The 6X DLG measured value for HDMLC changed from an average of ~0.8 mm in 2012 to ~0.4 mm in 2016. The measured DLG, or a slightly modified value, is typically used in Eclipse modeling. Currently Eclipse (V. 11 and 13) uses one DLG value per photon energy. Validation of the MLC modeling is typically performed by measuring fluence patterns, IMRT, and RA plans. Tests on two TBSTx showed that physical DLG <0.4 mm resulted in close agreement for fluence patterns and IMRT tests but resulted in a consistently higher ratio of measured to predicted dose for RA plans. A large value (~1.3 mm) in Eclipse was needed to accurately model RA plans. The physical gap on the two TBSTx were opened to a value around 0.5 mm. The Eclipse DLG was modeled with a value within <0.15 mm from the measured value for both TBSTx. The agreement for the fluence patterns, gap tests, IMRT and RA plans was evaluated. Results Using a larger modeled value in Eclipse (~1.3 mm) improved the agreement for most RA plans but resulted in dose discrepancies for fluence patterns and complex or small field IMRT plans. In one case, the chair fluence pattern typically used for commissioning tests resulted in a gamma analysis of 92% compared to the original 99% (2%-2mm; 10% threshold). Opening the HDMLC physical gap slightly to a value around 0.5 mm and entering a value in Eclipse closer to the measured value produced close agreement between measured and predicted dose for RA plans (Figs. 1 & 2). Dynamic fluence patterns and IMRT test plans maintained a similar agreement before and after the DLG physical changes.