In intensity modulated radiation treatment (IMRT) planning, the use of asymmetrically collimated fields that are placed on central axis or its off-set is mostly required. Output is the main topic discussed today for e...In intensity modulated radiation treatment (IMRT) planning, the use of asymmetrically collimated fields that are placed on central axis or its off-set is mostly required. Output is the main topic discussed today for extremely small and/or severe irregularly shaped fields. The air scatter data are involved directly or indirectly in obtaining the output. Despite the fact that extensive data have been published in many studies to provide a guide on the magnitude of output factor for clinical accelerators, there are very few data reviewed about output factor in-air or phantom for off-set fields. This study was aimed to investigate the impact of these conditions for small fields. This study was conducted in Elekta Synergy linear accelerator which produces 6 MV X-ray energy. The in-air output factor (Sc) has been measured by CC04 ion chamber with brass-alloy “build-up” cap and Dose-1 electrometer, and the total output (Scp) measurements were carried on at dose maximum depth in phantom by the same chamber and Thermoluminescence dosimeter (TLD) for 1 - 10 cm2 fields. The all measurements at center of isocenter and off-set fields at three directions (X2, Y1, Diagonal) were done. By decreasing field size from 10 to 2 cm2 at isocenter, the Sc value using CC04 was decreased to 5.4% and Scp using CC04 and TLD to 14.5% and 11% respectively. By increasing off-set value, the Sc and Scp values were increased in all directions comparing to central fields. The maximum increase was obtained in Y1 direction for Sc and Scp. TLD results for Scp is slightly higher than CC04. The dosimetric properties of small fields and their off-set should be evaluated and modelled appropriately in the treatment planning system to ensure accurate dose calculation in Intensity Modulated Radiation Treatment.展开更多
This study aimed to determine variations in tomotherapy beam outputs at multiple institutions. Measurements were obtained at 22 radiotherapy institutions. The first parameter was the absolute dose to water (Dfmsrw, Qm...This study aimed to determine variations in tomotherapy beam outputs at multiple institutions. Measurements were obtained at 22 radiotherapy institutions. The first parameter was the absolute dose to water (Dfmsrw, Qmsr) in the machine-specific reference field (fmsr), which indicated a static field in the tomotherapy reference conditions defined by the International Atomic Energy Agency (IAEA) study group. The second measured parameter was the difference between the measured and the planed doses in the intensity modulated radiotherapy (IMRT) verification plans, which were created using a solid phantom by the vendor during tomotherapy apparatus installation to adjust the beam output. The IMRT verification plan error at each institution was defined as the systematic error of the beam output;Dfmsrw, Qmsr was subsequently modified. The Dfmsrw, Qmsr values of four institutions with a modified energy fluence per ideal open time (EFIOT) were lower than the values at other institutions. The mean value of all institutions except those four was 0.994 ± 0.013 Gy (range: 0.974 Gy, 1.017 Gy). When the Dfmsrw, Qmsr value was corrected by the IMRT verification error, this variation decreased. In addition, the mean IMRT verification errors in the TomoDirectTM and TomoHelicalTM modes with the TomoEDGETM mode were 1.2% ± 0.8% (range: -0.6%, 1.8%) and 0.2% ± 0.5% (range: -0.6%, 0.9%), respectively (p展开更多
物理信息神经网络(physics-informed neural networks,PINN)由于嵌入了物理先验知识,可以在少量训练数据的情况下获得自动满足物理约束的代理模型,受到了智能科学计算领域的广泛关注.但是,PINN的离散时间模型(PINN-RK)无法同时近似多个...物理信息神经网络(physics-informed neural networks,PINN)由于嵌入了物理先验知识,可以在少量训练数据的情况下获得自动满足物理约束的代理模型,受到了智能科学计算领域的广泛关注.但是,PINN的离散时间模型(PINN-RK)无法同时近似多个物理量相互耦合的偏微分方程系统,限制了其处理复杂多物理场的能力.为了打破这一限制,文章提出了一种基于龙格库塔法的多输出物理信息神经网络(multi-output physics-informed neural networks based on the Runge-Kutta method,MO-PINN-RK),MO-PINN-RK模型在离散时间模型的基础上采用了并行输出的神经网络结构,通过将神经网络划分为多个子网络,建立了多个神经网络输出层.采用不同输出层近似不同物理量的方式,MO-PINN-RK模型不仅可以同时表征多个物理量,而且还能够实现求解偏微分方程系统的目的.另外,MO-PINN-RK克服了PINN离散时间模型仅适用于一维空间的局限性,将其应用范围扩展到了更为普遍的多维空间.为了验证MO-PINN-RK的有效性,文章对圆柱绕流问题进行了流场预测和参数辨识研究.测试结果表明,与PINN相比,MO-PINN-RK在流场预测问题中的准确性获得了提升,其精度至少提高了2倍,而在参数辨识问题中,MO-PINN-RK的相对误差降低了一个数量级.这凸显了MO-PINN-RK在流体动力学领域的卓越能力,为解决复杂问题提供了更准确、更有效的解决方案.展开更多
文摘In intensity modulated radiation treatment (IMRT) planning, the use of asymmetrically collimated fields that are placed on central axis or its off-set is mostly required. Output is the main topic discussed today for extremely small and/or severe irregularly shaped fields. The air scatter data are involved directly or indirectly in obtaining the output. Despite the fact that extensive data have been published in many studies to provide a guide on the magnitude of output factor for clinical accelerators, there are very few data reviewed about output factor in-air or phantom for off-set fields. This study was aimed to investigate the impact of these conditions for small fields. This study was conducted in Elekta Synergy linear accelerator which produces 6 MV X-ray energy. The in-air output factor (Sc) has been measured by CC04 ion chamber with brass-alloy “build-up” cap and Dose-1 electrometer, and the total output (Scp) measurements were carried on at dose maximum depth in phantom by the same chamber and Thermoluminescence dosimeter (TLD) for 1 - 10 cm2 fields. The all measurements at center of isocenter and off-set fields at three directions (X2, Y1, Diagonal) were done. By decreasing field size from 10 to 2 cm2 at isocenter, the Sc value using CC04 was decreased to 5.4% and Scp using CC04 and TLD to 14.5% and 11% respectively. By increasing off-set value, the Sc and Scp values were increased in all directions comparing to central fields. The maximum increase was obtained in Y1 direction for Sc and Scp. TLD results for Scp is slightly higher than CC04. The dosimetric properties of small fields and their off-set should be evaluated and modelled appropriately in the treatment planning system to ensure accurate dose calculation in Intensity Modulated Radiation Treatment.
文摘This study aimed to determine variations in tomotherapy beam outputs at multiple institutions. Measurements were obtained at 22 radiotherapy institutions. The first parameter was the absolute dose to water (Dfmsrw, Qmsr) in the machine-specific reference field (fmsr), which indicated a static field in the tomotherapy reference conditions defined by the International Atomic Energy Agency (IAEA) study group. The second measured parameter was the difference between the measured and the planed doses in the intensity modulated radiotherapy (IMRT) verification plans, which were created using a solid phantom by the vendor during tomotherapy apparatus installation to adjust the beam output. The IMRT verification plan error at each institution was defined as the systematic error of the beam output;Dfmsrw, Qmsr was subsequently modified. The Dfmsrw, Qmsr values of four institutions with a modified energy fluence per ideal open time (EFIOT) were lower than the values at other institutions. The mean value of all institutions except those four was 0.994 ± 0.013 Gy (range: 0.974 Gy, 1.017 Gy). When the Dfmsrw, Qmsr value was corrected by the IMRT verification error, this variation decreased. In addition, the mean IMRT verification errors in the TomoDirectTM and TomoHelicalTM modes with the TomoEDGETM mode were 1.2% ± 0.8% (range: -0.6%, 1.8%) and 0.2% ± 0.5% (range: -0.6%, 0.9%), respectively (p
文摘物理信息神经网络(physics-informed neural networks,PINN)由于嵌入了物理先验知识,可以在少量训练数据的情况下获得自动满足物理约束的代理模型,受到了智能科学计算领域的广泛关注.但是,PINN的离散时间模型(PINN-RK)无法同时近似多个物理量相互耦合的偏微分方程系统,限制了其处理复杂多物理场的能力.为了打破这一限制,文章提出了一种基于龙格库塔法的多输出物理信息神经网络(multi-output physics-informed neural networks based on the Runge-Kutta method,MO-PINN-RK),MO-PINN-RK模型在离散时间模型的基础上采用了并行输出的神经网络结构,通过将神经网络划分为多个子网络,建立了多个神经网络输出层.采用不同输出层近似不同物理量的方式,MO-PINN-RK模型不仅可以同时表征多个物理量,而且还能够实现求解偏微分方程系统的目的.另外,MO-PINN-RK克服了PINN离散时间模型仅适用于一维空间的局限性,将其应用范围扩展到了更为普遍的多维空间.为了验证MO-PINN-RK的有效性,文章对圆柱绕流问题进行了流场预测和参数辨识研究.测试结果表明,与PINN相比,MO-PINN-RK在流场预测问题中的准确性获得了提升,其精度至少提高了2倍,而在参数辨识问题中,MO-PINN-RK的相对误差降低了一个数量级.这凸显了MO-PINN-RK在流体动力学领域的卓越能力,为解决复杂问题提供了更准确、更有效的解决方案.
