Abstract：
BACKGROUND: MR-integrated proton therapy is under development. It consists of the unique challenge of integrating a proton pencil beam scanning (PBS) beam line nozzle with an magnetic resonance imaging (MRI) scanner. The magnetic interaction between these two components is deemed high risk as the MR images can be degraded if there is cross-talk during beam delivery and image acquisition.
OBJECTIVE: To create and benchmark a self-consistent proton PBS nozzle model for empowering the next stages of MR-integrated proton therapy development, namely exploring and de-risking complete integrated prototype system designs including magnetic shielding of the PBS nozzle.
METHODS: Magnetic field (COMSOL Multiphysics ${\\text{Multiphysics}}$ ) and radiation transport (Geant4) models of a proton PBS nozzle located at OncoRay (Dresden, Germany) were developed according to the manufacturers specifications. Geant4 simulations of the PBS process were performed by using magnetic field data generated by the COMSOL Multiphysics ${\\text{Multiphysics}}$ simulations. In total 315 spots were simulated which consisted of a 40 × 30 cm 2 $40\\times 30\\,{\\text{cm}}^{2}$ scan pattern with 5 cm spot spacings and for proton energies of 70, 100, 150, 200, and 220 MeV. Analysis of the simulated deflection at the beam isocenter plane was performed to determine the self-consistency of the model. The magnetic fringe field from a sub selection of 24 of the 315 spot simulations were directly compared with high precision magnetometer measurements. These focused on the maximum scanning setting of ± $\\pm$ 20 cm beam deflection as generated from the second scanning magnet in the PBS for a proton beam energy of 220 MeV. Locations along the beam line central axis (CAX) were measured at beam isocenter and downstream of 22, 47, 72, 97, and 122 cm. Horizontal off-axis positions were measured at 22 cm downstream of isocenter ( ± $\\pm$ 50, ± $\\pm$ 100, and ± $\\pm$ 150 cm from CAX).
RESULTS: The proton PBS simulations had good spatial agreement to the theoretical values in all 315 spots examined at the beam line isocenter plane (0-2.9 mm differences or within 1.5 % of the local spot deflection amount). Careful analysis of the experimental measurements were able to isolate the changes in magnetic fields due solely to the scanning magnet contribution, and showed 1.9 ± $\\pm$ 1.2 μ T $\\bf{\\mu} {\\text{T}}$ -9.4 ± $\\pm$ 1.2 μ T $\\bf{\\mu} {\\text{T}}$ changes over the range of measurement locations. Direct comparison with the equivalent simulations matched within the measurement apparatus and setup uncertainty in all but one measurement point.
CONCLUSIONS: For the first time a robust, accurate and self-consistent model of a proton PBS nozzle assembly has been created and successfully benchmarked for the purposes of advancing MR-integrated proton therapy research. The model will enable confidence in further simulation based work on fully integrated designs including MRI scanners and PBS nozzle magnetic shielding in order to de-risk and realize the full potential of MR-integrated proton therapy.
OBJECTIVE: To create and benchmark a self-consistent proton PBS nozzle model for empowering the next stages of MR-integrated proton therapy development, namely exploring and de-risking complete integrated prototype system designs including magnetic shielding of the PBS nozzle.
METHODS: Magnetic field (COMSOL Multiphysics ${\\text{Multiphysics}}$ ) and radiation transport (Geant4) models of a proton PBS nozzle located at OncoRay (Dresden, Germany) were developed according to the manufacturers specifications. Geant4 simulations of the PBS process were performed by using magnetic field data generated by the COMSOL Multiphysics ${\\text{Multiphysics}}$ simulations. In total 315 spots were simulated which consisted of a 40 × 30 cm 2 $40\\times 30\\,{\\text{cm}}^{2}$ scan pattern with 5 cm spot spacings and for proton energies of 70, 100, 150, 200, and 220 MeV. Analysis of the simulated deflection at the beam isocenter plane was performed to determine the self-consistency of the model. The magnetic fringe field from a sub selection of 24 of the 315 spot simulations were directly compared with high precision magnetometer measurements. These focused on the maximum scanning setting of ± $\\pm$ 20 cm beam deflection as generated from the second scanning magnet in the PBS for a proton beam energy of 220 MeV. Locations along the beam line central axis (CAX) were measured at beam isocenter and downstream of 22, 47, 72, 97, and 122 cm. Horizontal off-axis positions were measured at 22 cm downstream of isocenter ( ± $\\pm$ 50, ± $\\pm$ 100, and ± $\\pm$ 150 cm from CAX).
RESULTS: The proton PBS simulations had good spatial agreement to the theoretical values in all 315 spots examined at the beam line isocenter plane (0-2.9 mm differences or within 1.5 % of the local spot deflection amount). Careful analysis of the experimental measurements were able to isolate the changes in magnetic fields due solely to the scanning magnet contribution, and showed 1.9 ± $\\pm$ 1.2 μ T $\\bf{\\mu} {\\text{T}}$ -9.4 ± $\\pm$ 1.2 μ T $\\bf{\\mu} {\\text{T}}$ changes over the range of measurement locations. Direct comparison with the equivalent simulations matched within the measurement apparatus and setup uncertainty in all but one measurement point.
CONCLUSIONS: For the first time a robust, accurate and self-consistent model of a proton PBS nozzle assembly has been created and successfully benchmarked for the purposes of advancing MR-integrated proton therapy research. The model will enable confidence in further simulation based work on fully integrated designs including MRI scanners and PBS nozzle magnetic shielding in order to de-risk and realize the full potential of MR-integrated proton therapy.
摘要：
背景：MR整合质子治疗正在开发中。它包括将质子笔形束扫描（PBS）束线喷嘴与磁共振成像（MRI）扫描仪集成在一起的独特挑战。这两个部件之间的磁相互作用被认为是高风险的,因为如果在射束递送和图像采集期间存在串扰,则MR图像可能劣化。
目的：创建自洽的质子PBS喷嘴模型并对其进行基准测试，以增强MR整合质子治疗发展的下一阶段，即探索和去风险完整的集成原型系统设计，包括PBS喷嘴的磁屏蔽。
方法：位于OncoRay(德累斯顿,德国）是根据制造商的规格开发的。通过使用由COMSOLMultiphysics${\\\text{Multiphysics}}$模拟产生的磁场数据,对PBS工艺进行Geant4模拟。总共模拟了315个斑点，包括40×30cm2$40\\times30\\，{\\text{cm}}^{2}$扫描图案,光斑间距为5cm,质子能量为70、100、150、200和220MeV。对梁等中心平面的模拟挠度进行了分析，以确定模型的自洽性。将来自315个光斑模拟中的24个的子选择的磁场边缘场直接与高精度磁力计测量值进行比较。这些集中在从PBS中的第二扫描磁体产生的±$$20cm束偏转的最大扫描设置上，质子束能量为220MeV。在光束等中心和22、47、72、97和122cm的下游测量沿光束线中心轴(CAX)的位置。在等中心下游22厘米处测量水平离轴位置（距CAX±$$50，±$\\pm$100和±$\\pm$150厘米）。
结果:质子PBS模拟与在束线等中心平面处检查的所有315个斑点中的理论值具有良好的空间一致性(0-2.9mm差异或在局部斑点偏转量的1.5%内)。对实验测量的仔细分析能够隔离磁场的变化，仅由于扫描磁体的贡献，并显示1.9±$\\pm$1.2μT$\\bf{\\mu}{\\text{T}}$-9.4±$\\pm$1.2μT$\\bf{\\mu}{\\text{T}}$在测量位置范围内的变化。与测量设备内匹配的等效模拟进行直接比较，并在除一个测量点以外的所有测量点中进行设置不确定性。
结论：第一次，质子PBS喷嘴组件的准确和自洽的模型已经创建，并成功地进行了基准测试，以推进MR集成质子治疗研究。该模型将使人们对基于包括MRI扫描仪和PBS喷嘴磁屏蔽在内的完全集成设计的进一步仿真充满信心，以降低风险并实现MR集成质子治疗的全部潜力。
目的：创建自洽的质子PBS喷嘴模型并对其进行基准测试，以增强MR整合质子治疗发展的下一阶段，即探索和去风险完整的集成原型系统设计，包括PBS喷嘴的磁屏蔽。
方法：位于OncoRay(德累斯顿,德国）是根据制造商的规格开发的。通过使用由COMSOLMultiphysics${\\\text{Multiphysics}}$模拟产生的磁场数据,对PBS工艺进行Geant4模拟。总共模拟了315个斑点，包括40×30cm2$40\\times30\\，{\\text{cm}}^{2}$扫描图案,光斑间距为5cm,质子能量为70、100、150、200和220MeV。对梁等中心平面的模拟挠度进行了分析，以确定模型的自洽性。将来自315个光斑模拟中的24个的子选择的磁场边缘场直接与高精度磁力计测量值进行比较。这些集中在从PBS中的第二扫描磁体产生的±$$20cm束偏转的最大扫描设置上，质子束能量为220MeV。在光束等中心和22、47、72、97和122cm的下游测量沿光束线中心轴(CAX)的位置。在等中心下游22厘米处测量水平离轴位置（距CAX±$$50，±$\\pm$100和±$\\pm$150厘米）。
结果:质子PBS模拟与在束线等中心平面处检查的所有315个斑点中的理论值具有良好的空间一致性(0-2.9mm差异或在局部斑点偏转量的1.5%内)。对实验测量的仔细分析能够隔离磁场的变化，仅由于扫描磁体的贡献，并显示1.9±$\\pm$1.2μT$\\bf{\\mu}{\\text{T}}$-9.4±$\\pm$1.2μT$\\bf{\\mu}{\\text{T}}$在测量位置范围内的变化。与测量设备内匹配的等效模拟进行直接比较，并在除一个测量点以外的所有测量点中进行设置不确定性。
结论：第一次，质子PBS喷嘴组件的准确和自洽的模型已经创建，并成功地进行了基准测试，以推进MR集成质子治疗研究。该模型将使人们对基于包括MRI扫描仪和PBS喷嘴磁屏蔽在内的完全集成设计的进一步仿真充满信心，以降低风险并实现MR集成质子治疗的全部潜力。
