Purpose: To evaluate consistency in beam-data for a group of TrueBeam linacs.
Methods: Five TrueBeam linacs were installed between 2010 and 2018 at our institution. For each linac, beam-data was measured for input into TPS,including scanning data (PDD,crossline profile,diagonal profile) acquired using one of two scanning chamber types(0.04cc volume and 0.13cc volume)and output factors(0.13cc chamber volume).From the full dataset, an average dataset(‘consensus’) was generated for analysis, and variation of each machine dataset was evaluated relative to consensus data.For scanning data, field sizes larger than 4x4cm2were included for analysis, ranging from 6x6cm² to 40x40cm²(6 field sizes in total).For output-factor data, field sizes 3x3cm² and larger were included for analysis. The data was evaluated as follows:(1) output factors – relative-st.dev. is reported,(2)PDD - average dose difference magnitude and overall max dose difference for depths beyond dmax are reported,(3)crossline/diagonal profiles – average dose difference magnitude and overall max dose difference within the inner field (defined as inner 80% of field).
Results: For output-factor comparison,the average relative-st.dev. was 0.15% and largest relative-st.dev. was 0.37%. For PDD measurements, range in average dose difference magnitude was [0.18%,0.26%] and range in overall max dose difference was[0.25%,0.83%]. For crossline profiles (field sizes of 30x30cm² and smaller), the range in average dose difference magnitude was [0.06%,0.19%],and range in overall max dose difference was[0.52%,0.65%]. For 40x40cm²2 field size, overall max dose differences were slightly larger, with range of [0.64%,1.17%].For diagonal profiles, range in average dose difference magnitude was [0.19%,0.37%] and range in overall max dose difference was[0.55%,1.29%].
Conclusion: For field sizes ranging from 6x6cm² to 30x30cm², our results indicate a level of agreement well within 1% for all measurements.For largest field size(40x40cm²),our results indicate a level of agreement within 1.5% for all measurements. Overall,our results suggest that average beam-data set may be used to model the broad-beam-data for TrueBeam linacs.
Not Applicable / None Entered.