Positions of radiation isocenter and the couch rotation center established by Winston-Lutz and optical measurements

PURPOSE: To compare x-ray and optical imaging methods for measuring the relative position of radiation isocenter and couch rotation center. To show the impact of radiation isocenter size and target motion on the margins for target contours. METHODS: Winston-Lutz measurements are made using EPID images. Image analysis was done with public domain software, ImageJ, and spreadsheets written in Microsoft Excel. A comparison between the center of a high density test object and center of the MLC collimated beam is used to judge the relative position of the radiation isocenter in space for gantry and couch rotation. Additionally, motion of the target with couch rotation is determined with an optical imaging system. Five different accelerators, two TrueBeams, a Trilogy, and two VersaHDs, were assessed by Winston-Lutz and optical methods. RESULTS: The shift in the radiation isocenter with gantry rotation is found to be a tri-axial ellipsoid. Shifts in the target position with respect to radiation isocenter with couch rotation were between 0.4 and 0.6 mm. The Winston-Lutz and optical method determination of couch rotation center agreed within measurement uncertainty. CONCLUSIONS: Image analysis yields precise data on linear accelerator radiation isocenter and rotation centers of the couch. The Winston-Lutz and optical methods agreed within measurement uncertainty.