Room: AAPM ePoster Library
Purpose: Standard practice is to avoid beam entry through rails, although VMAT treatments make rail avoidance for a subset of patients unrealistic; moreover, setup error may result in fields entering through rails.
Eclipse rail models use a simplified drawing as opposed to a CT image or precise technical drawing including the full rail structure. We present measurements which quantify the consequences of simplified rail models for a number of nominal fields and arcs delivered through rails.
Methods: An ArcCheck® detector was aligned at isocenter, data were acquired with and without rails present using 6MV, 100MU, 10x10cm fields at a number of posterior angles, as well as angles of 0° and 90° to investigate backscatter and provide a reference measurement. Partial arcs through the posterior rail section were also analyzed. Gamma analysis pass rates were determined at 3%/3mm, 2%/2mm and 1%/1mm using a threshold of 10%. Measurements were acquired using a Varian TrueBeam® and kVue™ Couch Top.
Results: Results for gantry angles of 90° and 0° were identical, suggesting no effect of rail model on backscatter accuracy. Gamma passing rates were lower for all posterior fields delivered through rails as opposed to those delivered without the rail. Dosimetric errors, both hot and cold, were observed in regions of beam entry and exit. All distributions showed point errors in exceeding 5%. Passing rates were higher, relative to static fields, for the 10x10cm arc, but comparable for the 5x5cm arc. When limiting the analysis to the region measuring entrance dose, gamma pass rates dropped to 83% and 57% for the 10x10 and 5x5 arcs, respectively.
Conclusion: Eclipse rail models are largely accurate for large fields and do not appear to introduce significant error at conventional gamma analysis levels. At the 1%/1mm level there is considerable dosimetric inaccuracy impacting both static and arc fields.