END-TO-END METHOD FOR INTEGRATED ASSESSMENT OF DIFFERENT IGRT CBCT, OSMS, CALYPSO AND IMAGES PLANS KV AND MV Methods:

This paper aims to establish a method for simultaneous evaluation of different IGRT systems present in TrueBeam STx / Edge. Thus, Surface Guidance methods (OSMS), Cone Beam CT (CBCT), Planar Images MV-KV and Calypso are tested and compared on a single phantom, allowing the determination of positioning accuracy relative to the radiative isocenter. The sensitivity of each system to detect submillimeter displacements was also evaluated. The phantom consists of a solid water cube with three beacons and an 8 mm diameter metal sphere. From it, a CT and a RT plan with square fields centered on the sphere were made – simulating a Winston-Lutz (WL) type test. The experiment was divided into 2 parts: (i) the phantom was positioned with CBCT (0.5×0.5×1.0 mm³ resolution, 6D and 3D manual and automatic recording methods). From this point, the displacement indicated by the other methods was recorded: OSMS, Calypso and kV / MV pairs (2D-3D method). The actual displacement from the radioactive isocenter was found through the planned field gateways (WL test) – this step was repeated 10 times. (ii) Then table displacements (minimum pitch: 0.1mm, total of 25 positions) were applied along each axis (lateral, longitudinal and vertical) and the displacement indications detected by each IGRT method were recorded. The following mean vector deviations from radiative isocenter were found for CBCT (manual fusion, 6D and 3D), OSMS, Calypso, pairs (MV-MV, KV-KV and MV-KV) respectively: 0.43 ± 0 .09; 0.48 ± 0.09; 0.47 ± 0.09; 0.69 ± 0.10; 0.92 ± 0.35; 0.64 ± 0.10; 0.61 ± 0.10 and 0.62 ± 0.10 mm. The sensitivity results in detecting table displacements (in any direction) of OSMS (N = 75), CBCT (N = 51), planar KV / MV (N = 51) and Calypso (N = 75) systems were, respectively: 0.04 ± 0.05; 0.08 ± 0.08; 0.24 ± 0.43 and 0.07 ± 0.12mm. No significant differences were observed along a specific direction. With these results, it is possible to obtain confidence intervals of less than 1 mm to verify the calibration of each IGRT system. Therefore, the presented method is satisfactory for periodic QA, and allows simultaneous verification of all systems.