Accelerated and standardized commissioning of a robotic RadioSurgery system using a hybrid golden beam data model

E. Pantelis, A. Moutsatsos, L. Sideri, P. Archontakis, Anastasia Stergioula

Abstract Background The CyberKnife™ system requires extensive beam data acquisition across multiple collimation systems. The CyberComm™ framework integrates a linac output tuning procedure with a hybrid Golden Beam Data (GBD) model to accelerate and standardize CyberKnife commissioning while maintaining dosimetric accuracy. Purpose In this study, CyberComm was employed for the commissioning of a CyberKnife model S7 system and evaluated with regard to its efficiency and dosimetric accuracy. Methods Following the CyberComm framework, commissioning was separated into two stages. First, the linac beam was tuned by acquiring percentage depth dose (PDD) and off‐center ratio (OCR) measurements for the 60 mm fixed collimator (Cone‐60) and the primary (uncollimated) beam. These measurements were compared with corresponding reference datasets using one‐dimensional gamma index (GI) acceptance criteria of 1% (local)/1 mm for PDD, and 0.7% (global)/1 mm for OCR profiles. The process was repeated until GI passing rates exceeding 95% for all measurements. In the second stage, full beam commissioning measurements were performed for the smaller field sizes, while for the larger ones, the corresponding GBD values were validated through spot‐checks. Dosimetry measurements were conducted using a PTW‐60019 microdiamond detector. Dosimetric accuracy of the commissioned beam models was further evaluated using single beam and patient‐specific quality assurance (PSQA) measurements with the SRS MapCHECK™ detector. Results Using vendor‐provided linac pre‐sets, the initial Cone‐60 PDD yielded a GI passing rate of 61%, while OCR profiles achieved 100%. After tuning the PFN‐HVPS voltage and steering coil currents, Cone‐60 PDD GI passing rate improved to 98%, with both primary beam and collimated OCR profiles achieving 100%. Spot‐check validation demonstrated excellent agreement between measured beam data and the GBD model, with 100% GI passing rates for TPR and OCR profiles across all evaluated field sizes. Output factors agreed within 0.4% with vendor reference data for the majority of fields, with larger deviations observed only for the smallest apertures. Single beam and PSQA irradiations showed increased dosimetric accuracy, with GI passing rates of 100% and ≥ 97%, respectively, using 2%/1 mm GI acceptance criteria. Complete commissioning of all collimation systems was achieved within 10 days. Conclusions CyberComm‐assisted commissioning demonstrated accurate beam model generation for the CyberKnife system with a reduction in commissioning time relative to conventional protocols. Although these results are based on a single‐center installation, they support the clinical feasibility of this workflow for CyberKnife commissioning.

Also available via: Journal of Applied Clinical Medical Physics