Author: Mylinh Dang, Indra J. Das, George X. Ding, Ahtesham Ullah U Khan, Andrew J. White 👨🔬
Affiliation: Northwestern Memorial Hospital, Northwestern University Feinberg School of Medicine, Northwestern Medicine, Vanderbilt University Medical Center 🌍
Purpose: The large variability in field output factor (FOF) observed among various studies is speculated to be due to uncertainty in jaw reproducibility that is linked directly with source occlusion. Most machines’ field size is defined in centimeters with accuracy of 0.1 cm. The reproducibility of jaw positions and associated uncertainty in measured FOF along with Mone Carlo (MC) simulations is studied.
Methods: Using a scanning water tank and microdetector (microSilicon), beam profiles for 6 MV at 95 cm SSD at 5 cm depth were acquired to the highest permissible accuracy (0.1 -0.2 mm) in X and Y planes for 5-, 8-, 10-, 30- and 100-mm square fields from newly commissioned Varian TrueBeam accelerators. Fields were scanned 10 times by opening and closing the jaw sizes to desired fields for profiles and associated FOF were measured.
Results: The reproducibility of jaw position is found to be highly accurate as seen by 10 profiles, but when numerically analyzed, it was found to be in the fraction of 0.1-0.2 mm (<1%). The magnitude however differed among machines up to <2% based on standard deviation in 10 scanned data. The jaw variability did not translate to the associate dose variability among machines and was found to be <0.2% for 5 mm field. Using MC simulation, it is shown that FOF changes 2%/0.1mm. Hence it is explainable that up to 5% changes in FOF could be due to jaw variability and associated source occlusion.
Conclusion: Modern machines reproduce jaw position (<0.2 mm) with minimum changes in FOF. The uncertainties in FOF for small fields is probably due to source occlusion associated with variability in jaw position. MC simulation could provide data accurately and can explain large variation in FOF with jaw variation of 0.1-0.5 mm.