[Impact of Imaging and Reconstruction Parameters for Cone-beam Computed Tomography on Three-dimensional Star Shot...

Previous studies have not demonstrated the influence of various imaging acquisition and reconstruction conditions on three-dimensional star shot analysis using an X-ray-based gel dosimeter. Therefore, we investigated the effects of image acquisition and reconstruction conditions on the method in this study and evaluated the robustness of the method. Firstly, we set a reference condition (80 kVp, 1080 mAs, full fan, full trajectory, 1800 projection number, iterative reconstruction) and nine other comparison conditions (80-125 kVp, 540-2160 mAs, half or full fan, full or half trajectory, 420-5400 projection number and iterative reconstruction or filtered back projection). Cone-beam computed tomography scans were performed under each condition before and after irradiation of the single gel dosimeter, and the position of the radiation isocenter relative to the image isocenter, beam isocentricity, and contrast-to-noise ratio (CNR) was measured. The differences compared to the reference were as follows: for isocenter distance, 0.0±0.1 mm on the X-axis, 0.0±0.2 mm on the Y-axis, and 0.1±0.3 mm on the Z-axis. The beam isocentricity was 0.0±0.2 mm. The CNR ranged from a maximum of 5.8 to a minimum of 2.6 in median comparisons. In contrast, compared to the results under the reference condition, the results under other conditions matched within a range of ±0.25 mm. This investigation showed that analysis is feasible within an uncertainty of 0.25 mm under imaging conditions where CNR is 3 or higher.