Room: Exhibit Hall | Forum 5
Purpose: To assess the accuracy and reproducibility of optical surface tracking (OSMS) compared to kV and MV on-board imaging modalities as part of linac commissioning for three Varian machines (two TrueBeams and one Edge).
Methods: Simple treatment plans were created in Eclipse for an anthropomorphic pelvis phantom with a dark skin tone. Imaging accuracy was tested by applying known translational and rotational shifts and comparing these to the shifts determined by each imaging technique (kV and MV planar, kV CBCT, OSMS). Reproducibility was determined by repeatedly using the imaging systems to shift the phantom from the user origin to the treatment isocenter and by repeatedly applying and correcting for known shifts.
Results: Translational reproducibility was within 0.8mm for kV and MV planar techniques and 0.4mm for CBCT. Accuracy was achieved within 0.8mm and 0.6 degrees for all techniques with the exception of the Image Gently CBCT technique, which agreed within 1.9mm, and the half-trajectory Head technique for one of the TrueBeams, which agreed within 1.1mm. OSMS accuracy was noticeably worse for the phantom used; translational and rotational shifts agreed with known values within 1.4mm and 0.7 degrees, respectively, and differed from CBCT by up to 2.1mm.
Conclusion: On-board imaging systems for all three machines performed well for the anthropomorphic phantom used. Some imaging techniques that showed reduced accuracy were not clinically applicable for the phantom used (e.g., â€œImage Gentlyâ€? CBCT for an adult-sized pelvis phantom), and their use for patient positioning should be carefully considered according to patient size and the site to be imaged. OSMS was less accurate and less reproducible for the phantom studied compared to other imaging modalities, and should be used in conjunction with other techniques. Special caution should be used when relying on OSMS for patients with darker skin tones.