Purpose: Single-isocenter multitarget radiosurgery (SIMS-RS) is a treatment option for multiple brain metastases. Precise alignment of rotational axes of the linear accelerator is crucial and must be checked frequently. The drawback of current quality assurance (QA) procedures is that targeting accuracy is only evaluated at isocenter. This study introduces a hidden target end-to-end (E2E) test allowing to check targeting accuracy for off-axis targets.
Methods: The E2E follows our current clinical workflow. Targets are three BBs placed inside the Rando head phantom. The procedure is similar to the Winston and Lutz isocentricity test (WL-test). A 10cmx10cm field size allows that the implanted BBs were visible in the recorded portal images. For evaluation, an in-hose tool for image analysis and 3D reconstruction of target positions was developed. The algorithm allows reconstruction of multiple BB positions in a non-homogeneous phantom. In addition, a multi-view 3D reconstruction is used to automatically identify the BBs and to calculate its 3D positions. The automated analysis consists of a classic WL-test for the central BB as well as a comprehensive 3D analysis of all BBs.
Results: The median distance for the center BB to the CAX was 0.46mm. The reconstructed size of the isocenter was 0.36mm for collimator/gantry combinations and 1.0mm for couch rotations. This is in agreement with an independent WL-test. For the off-axis BBs with an displacement of 5.6cm and 7.0cm the reconstructed center size was 0.49mm and 0.72mm, respectively. Couch rotation led to an uncertainty 1.05mm for the proximal and 1.26mm for the distal BB. A 3D point registration showed that the total target accuracy was 0.43mm.
Conclusion: The proposed E2E test in combination with the developed processing algorithm allows a streamlined assessment of SIMS-RT targeting accuracy and can be used as an additional QA step for those complex treatments.