Purpose: Intrafraction and interfraction motion are important challenges to address during radiotherapy treatment. To monitor these motions, image guidance methods have been developed to monitor the target intrafraction motion and the effects of interfraction anatomical changes on dose distributions. AAPM Task Groups 76 and 132 recommend before the clinical implementation of these systems, end-to-end validation should be performed. The goal of this work was to develop an anthropomorphic deformable phantom system which incorporates 3D dosimetry to validate image guidance systems.
Methods: A deformable abdominal phantom driven by a plunger was developed using plastisol polyvinyl-chloride (PVCP). Representative organs were fabricated with 3D printed molds of patient anatomy and material properties were tuned to match realistic anatomical CT values. The addition of graphite ultrasonic scatter resulted in realistic ultrasound target images for tracking. A cavity in the liver section of the phantom was created to hold a deformable 3D polymer gel dosimeter canister. Benchmarking of the phantom was performed by testing the repeatability of deformation and the accuracy of the 3D dosimetry.
Results: Fiducial marker positions in the phantom were repeatable within 0.1 mm on average in each direction over 15 deformations. Dice coefficients comparing the dosimeter position over all deformations were above 0.98. The gel dosimeter matched an ion chamber dose measurement within 3% during liver SBRT treatment delivery. Î³-analyses comparing the gel dose distribution to the TPS planned distribution and EBT3 film measurements resulted in average pass rates of 99.2% and 90.1% (3%/5 mm), and 99.6% and 93.4% (5%/5 mm), respectively.
Conclusion: This deformable phantom system has been shown to be repeatable and provide accurate 3D dose distributions. This phantom shows the potential to provide a platform for the testing of a variety of intra- and interfraction image guidance systems before clinical implementation.