Purpose: Multiple leaf collimator (MLC) is used as part of proton therapy pencil beam scanning system to provide better sparing to organs at risks. MLC position accuracy is very important to ensure accuracy of dose delivered compared to the planned dose. This study proposed and validated a method to use a pseudo Picket Fence and Open Field to identify MLC position errors.
Methods: Picket Fence and Open Field plans were generated in RayStation treatment planning system with proton energies at 45.85 MeV and 25.84 MeV, correspondingly. The spot size and penumbra are relatively large and can be used to decouple the positioning accuracy of leaf and spot. Images were acquired on an IBA Lynx PT imager with a 30 cm x 30 cm imager size and 0.5 mm resolution. Both data sets included one image acquisition without any intentionally induced MLC position error, and one or more acquisitions with induced error. For acquisitions with induced error, the Picket Fence plan was modified to increase the width of select pickets by 1 mm or 2 mm. For the Open Field, edges were decreased by 1mm, 2mm, and 5mm. Analysis was performed using custom python scripts.
Results: Picket Fence images consistently detected changes in the picket widths of the modified pickets when compared to the non-modified dataset. Changes to MLC positions of both 1mm and 2mm were detectable across images slice profiles, including the direction the MLC position adjustment. Open Field analysis was able to detect fine errors as low as 0.5mm on both field size and penumbra size as block margins were added in 0.5 mm increments.
Conclusion: This study show that both Picket Fence position, Field size and penumbra size are sensitive to MLC position errors. Picket Pence and Open field methods can provide quick and efficient tests to MLCs on a compact proton pencil beam scanning system.