ESTRO 38 Abstract book

S949 ESTRO 38

other SRS delivery systems such as TB. The test results are reproducible and consistent. The spatial accuracy of the CK’s 6DSkull tracking with MLC is better than TB’s kV-CBCT verification and 6DoF couch, but the difference is small. These results confirm that both systems are appropriate for safe SRS deliveries. EP-1758 A validation and criticality assessment of imaging dose calculation discrepancies of Halcyon MVCBCT Y. Zhang 1 , Y. Huang 1 , M. Wang 1 , R. Wang 1 , W. Wang 1 , H. Wu 1 1 Key Laboratory of Carcinogenesis and Translational Research Ministry of Education/Beijing- Department of Radiation Oncology- Peking University Cancer Hospital & Institute, Department of Radiotherapy, Beijing, China Purpose or Objective By using identical 6MV flattening-filter-free photons for both imaging and treatment on Varian Halcyon system, the total dose of the two can be incorporated automatically in Eclipse Treatment Planning system, which, however, uses treatment isocenter to calculate the imaging dose. This work aims to assess the criticality of the calculation deviations induced by fractional image-guided-couch- shifts for Halcyon MV cone beam CT (CBCT) imaging dose, based on real patient data. Material and Methods Eclipse-calculated imaging dose was first validated using a calibrated A1SL chamber in the ‘Cheese Phantom’ by measurement. Then, the actual imaging dose (D act ) for 18 historical patients of various sites were recalculated based on 513 MV CBCT-guided-couch-shift data, and compared with the reference computations based on treatment isocentre (D ref ). 'High Quality' and 'Low Dose' MV CBCT modes were applied in 382 and 131 fractions, delivering 10 MU and 5 MU per scan respectively. Patient- and plan- specific dose from treatment fields was integrated with D act and D ref respectively for comparison (Total Dose). More efficient computation was achieved by the automation using Eclipse Scripting Application Programming Interface. Results The average unsigned relative disagreements between the measured and Eclipse-calculated dose in Cheese phantom were less than 1.23%, hence the following studies were based on Eclipse simulations. Figure 1 displays the couch shifts on the lateral (a), vertical (b) and longitudinal (c) directions of 18 patients as guided by 513 MV CBCT scans. The bottom and top of the boxes indicate the first and third quartiles. The bands inside the boxes show the second quartile (the median). The whiskers display the lowest/highest datum still within 1.5 interquartile ranges of the lower/upper quartiles. The circles indicate the outliers.

As shown in table 1, the Gamma passing rates of the accumulated imaging dose were all better than 95.08% based on 3mm/3%/global/10% threshold criteria, and the mean±1SD was 97.54%±1.42%. The accumulated errors of minimum imaging dose to PTV were no larger than -14.38 cGy, which were reduced to -0.82 cGy after the heterogeneous treatment does was overlaid. The mean relative discrepancies of PTV minimum dose were -0.61 cGy (-0.71%) and -0.00 (0.00%), before and after incorporating the treatment dose respectively.

Conclusion The Eclipse-calculated Halcyon MV CBCT dose was validated by measurement. Although the isocentre displacement-induced imaging dose calculation discrepancies for Halcyon MV CBCT were partially cancelled out by couch shifts of various directions and distances, especially after the incorporation of