Efficient quality assurance method with automated data acquisition of a single phantom setup to determine radiation and imaging isocenter congruence

We developed a quality assurance (QA) method to determine the isocenter congruence of Optical Surface Monitoring System (OSMS, Varian, CA, USA), kilovoltage (kV), and megavoltage (MV) imaging, and the radiation isocenter using a single setup of the OSMS phantom for frameless Stereotactic Radiosurgery (SRS) treatment. After aligning the phantom to the OSMS isocenter, a cone-beam computed tomography (CBCT) of the phantom was acquired and registered to a computed tomography (CT) scan of the phantom to determine the CBCT isocenter. Without moving the phantom, MV and kV images were simultaneously acquired at four gantry angles to localize MV and kV isocenters. Then, Winston-Lutz (W-L) test images of the central BB in the phantom were acquired to analyze the radiation isocenter. The gantry and couch were automatically controlled using the TrueBeam Developer Mode during MV, kV, and W-L image acquisition. All the images were acquired weekly for 17 weeks to track the congruence of all the imaging modalities' isocenter in six-dimensional (6D) translations and rotations, and the radiation isocenter in three-dimensional (3D) translations. The shifts of isocenters of all imaging modalities and the radiation isocenter from the OSMS isocenter were within 0.2 mm and 0.2° on average over 17 weeks. The maximum discrepancy between OSMS and other imaging modalities or radiation isocenters was 0.8 mm and 0.3°. However, systematic shifts of radiation isocenter anteriorly and laterally relative to the OSMS isocenter were observed. The measured discrepancies were consistent from week-to-week except for two weeks when the isocenter discrepancies of 0.8 mm were noted due to drifts of the OSMS isocenter. Once recalibration was performed on OSMS, the discrepancy was reduced to 0.3 mm and 0.2°.By performing the proposed QA on a weekly basis, the isocenter congruencies of multiple imaging systems and radiation isocenter were validated for a linear accelerator.