Scintillation Imaging-Based 3D Printed Phantom for Robotic-Arm Linac AQA: Towards Rapid and Reusable Delivery Verification
Abstract
Purpose
This study demonstrates a robust scintillation imaging-based solution to robotic-arm linac automated quality assurance (AQA), including the translational targeting error verification that combines tracking and delivery discrepancies.
Methods
A 3D-printed phantom comprising an acrylic cube and an L-shaped transparent PMMA component was mounted on a custom-designed platform. The acrylic cube contains a tungsten ball at its geometric center. scintillating sheets were fixed onto the two orthogonal beam-exit surfaces of the acrylic cube component to provide high-contrast optical images, which were then reflected via a mirror and captured by a complementary metal-oxide semiconductor camera for AQA beam analysis. AQA tests for different collimators were performed and compared against EBT3 film data, with a focus on the centroid offset in each orientation and the overall radial error in the Patient Plane Coordinates. The sensitivity to changes in output and short-term stability were also assessed. A user-friendly software application was developed to enable near-real-time data analysis in the clinical setting.
Results
Centroid offsets measured by the optical system agreed closely with EBT3 film data, with differences of less than 0.67 mm in all orientations. For fixed cone collimators, the mean radial error was 0.37 ± 0.11 mm, and for MLC collimators it was 0.44 ± 0.15 mm, with both values remaining within the 0.95 mm tolerance. The dose response for the scintillation emission was linear, exhibiting excellent linearity (R² = 0.99) across a broad range of monitor units. Over an eight-day measurement period, the system demonstrated good reproducibility, with radial error contained within a consistent range.
Conclusion
The phantom provides a rapid, reusable solution for robotic‑arm linac AQA, directly measuring beam placement through optical imaging. Combined with dedicated analysis software, it offers a clinically deployable alternative to film‑based verification.