Author: Min-Sig Hwang, Danny K. Lee, Daniel C. Pavord, Kyung Lim Yun 👨🔬
Affiliation: Allegheny Health Network 🌍
Purpose: Ensuring the quality of treatment plans through patient-specific pre-treatment quality assurance (QA) is essential. However, the use of physical phantom-based QA devices is not feasible for online adaptive planning, and single-point dose verification may present uncertainties due to target heterogeneity. This study seeks to implement independent three-dimensional dose verification using RayStation® (Version 9B, RaySearch; Stockholm, Sweden) for online MR-Linac adaptive radiotherapy.
Methods: A virtual dose verification system was clinically implemented as follows: (1) A beam model of MR-Linac 7MV FFF was built on RayStation® using Monte Carlo® calculated 3D dose profiles from MRL Monaco. Asymmetric beam profiles due to a 1.5T magnetic field were used for the RayStation® model. (2) The model was validated with square and rectangular field profiles at a 10cm depth. (3) Clinical IMRT plans were validated using in-house gamma analysis software. (4) Clinical Monaco plans were imported to RayStation®, and doses from both systems were compared. (5) The mean PTV dose between RayStation® and Monaco was evaluated for clinical use.
Results: The RayStation® beam model of an MR-Linac 7MV FFF beam demonstrated an average 95% (2%/2mm) and 98% (3%/3mm) 3D gamma passing rates in simple fields, and an average 95% (3%/3mm) in 12 clinical plans, including liver, pancreas, node, prostate, rectum, and adrenal. The processing time for virtual QA was approximately 3 minutes from Monaco plan export to RayStation dose calculation and evaluation.
Conclusion: This study implemented a virtual QA system in RayStation® for independent dose verification, proving its effectiveness during online MR-guided adaptive radiotherapy with a reasonable time extension.