Author: Tsuicheng D. Chiu, Weiguo Lu, Aaron Thomlinson, You Zhang 👨🔬
Affiliation: Department of Radiation Oncology, UT Southwestern Medical Center, Medical Artificial Intelligence and Automation (MAIA) Lab, Department of Radiation Oncology, UT Southwestern Medical Center, University of Texas Southwestern Medical Center 🌍
Purpose: To develop and validate an MR-compatible anthropomorphic motion phantom to assess liver motion, real-time dosimetry, and gating system performance under controlled and reproducible respiratory conditions, addressing critical measurement/validation needs of gated MR-guided Adaptive Radiotherapy on a realistic phantom with complicated deformable motion.
Methods: An MR-compatible anthropomorphic motion phantom, scaled to 70% of nominal adult size, was developed to simulate realistic respiration-driven liver deformable motion and evaluate gating techniques. The phantom was equipped with a pneumatic pumping system to control lung inspiration levels and associated downstream motion which includes the liver, with an Arduino-based system managing digital control and data recording. Lung inflation and deflation were achieved using two 40 L/min vacuum pumps, four 12-VDC solenoid valves, and an H-Bridge motor controller. The modular liver component allowed for interchangeable liver shapes with and without tumor targets. Four scintillator dosimeters were embedded via brachytherapy catheters to enable real-time dosimetry. A pressure transducer monitored lung pressure, and logged in real-time. The phantom was scanned using an Elekta Unity MR-LINAC with T2 and Hass motion sequences to evaluate its motion capabilities.
Results: Experiments were conducted at excitation levels of 100%, 80%, 40%, and 20%, corresponding to average lung air pressures of 0.087, 0.071, 0.047, and 0.021 psi, respectively. Liver deformations were measured along Cantlie’s line and across four sessions in the coronal plane, as well as anterior, medial, and posterior locations of the sagittal plane, along with additional evaluations of liver diagonal changes. At 100% excitation, approximately 9.28 mm of liver deformation was observed along Cantlie’s line in the coronal plane.
Conclusion: This study highlights the successful development and application of an advanced anthropomorphic motion phantom capable of simulating realistic liver motion and enabling precise imaging/dosimetry measurements for MR sequence development and MR-LINAC gating validation.