Purpose: To verify the feasibility of using a single beam model for SBRT treatments across multiple linear accelerators through patient-specific QA measurements of representative SBRT plans.
Methods: Ten previously treated SBRT plans for various sites (lung, brain, spine, pancreas, adrenal gland, iliac lymph node) were created using a single unified beam model(AAA v.15.6) and 6FFF mode in Eclipse TPS. Input beam data (PDD,Crossline Profiles,Output Factors) used for the model were based on institutional consensus data, which used average measured data from multiple TrueBeam machines for each data type. The measured DLG and MLC transmission agreed within 0.1 mm and 0.01%(absolute), respectively, for the two machines analyzed. To test the beam model, plans were mapped to an acrylic phantom for point dose measurement using micro-ionization chamber(PTW PinPoint) and planar dose measurement(Gafchromic EBT3 film) on two linacs, with both dose measurements performed at isocenter. The same plan was delivered on each machine, and the plan dose was compared to measured dose using point dose difference, Gamma analysis, and line profile analysis.
Results: For point dose measurements, the average±st.dev. percent difference was 0.36±1.60% for Machine1 and 2.07%±1.47% for Machine2. All point dose measurements were within 4% of planned dose values. For film measurements, Gamma(3%,2mm) passing rates for both machines were in excess of 98.5% for all plans. Additionally, line profile analysis was performed to compare planar dose measurements to calculated plan dose using unified beam model. Line profile analysis indicated agreement in high dose region, penumbra, and low-dose tail for each machine relative to the planned dose.
Conclusion: Our results for point and planar dose indicate an acceptable level of agreement between measured dose and plan dose for each machine. We conclude that using a single unified beam model for SBRT plan delivery is a feasible option for clinics with multiple TrueBeam machines.