Purpose: In this study, a novel Cherenkov imaging-based isocenter coincidence verification method is presented as a clinically deployable, integrated solution for MR-Linac daily quality assurance (QA).
Methods: An acrylic water-filled cylindrical phantom was designed containing a custom-machined plastic conical structure fixed in place with 3D-printed spacers. Cherenkov emission from the cone was highly scattered, providing optical contrast from the surrounding water. The plastic is also visible with high contrast on low-field MR imaging, and the conical structure was etched with fiducials, which were visible on Cherenkov images. The phantom was aligned using co-registration of MR the images to baseline. A star shot plan was developed and delivered to the phantom to provide x-z isocenter localization relative fiducials on the phantom, while a four-field thin sheet plan yielding a ring-shaped Cherenkov emission distribution, which encodes isocenter localization along the y-axis. A custom analysis program with a graphical user interface was developed in MATLAB to analyze results in near-real-time.
Results: Strong linearity was observed (R² > 0.99) between longitudinal position and optical ring diameter using calibration data, and error was low (RMSE = 0.184 mm). Aa minimum circle radius of 0.34 mm was measured during the star shot analysis. Initial isocenter coincidence measurements in the lateral, longitudinal, and vertical directions were -0.61 mm, 0.55 mm, and -0.14 mm respectively, yielding a 3D coincidence of 0.83 mm, below the 2 mm tolerance defined in TG-142. Longitudinal analysis showed an average coincidence of 1.5 mm ± 0.4 mm over 8 weeks of daily use.
Conclusion: This method represents a robust and efficient method for MR-Linac isocenter coincidence verification which can be realistically deployed in the clinic for daily measurement. Longitudinal use data from our clinic suggests this method is a viable alternative to measurements which use reduced datasets or heavy time resources.
Funding Support, Disclosures, and Conflict of Interest: This work was funded by NIH grant R01EB023909. BP is the president and PB is the principal scientist at DoseOptics LLC, which supplied cameras and software for this study. DA receives consulting fees from DoseOptics outside the context of this work.