Zhang F, Qin Y, Segars W, Yin F, Cai J
Duke Medical Physics Program, Durham, NC.
Duke University Medical Center, Durham, NC.
Med Phys. 2012 Jun;39(6Part20):3859. doi: 10.1118/1.4735756.
At present, techniques of 4D dose verification in radiation therapy for lung cancers associated with respiratory motion are very limited, due to either lack of patient anatomy or breathing information. The aim of this study is to develop a 4D dose verification technique based on XCAT phantom which incorporates both patient anatomy and respiratory mechanics.
The 4D-XCAT phantom which possesses patient specific anatomy and allows input of respiratory parameters was generated using an integrated Matlab program. 10-phase 4DCT, MIP, and AIP images generated from 4D-XCAT were used for designing a 3D conformal treatment plan. Real-time dose delivery was simulated by calculating the deposited dose in each phase of 4D-XCAT for each beam with fixed individual planning dose. The 4D delivered dose was determined by accumulating dose deposition of all beams in all phases using deformable image registration implemented in VelocityAl software. The planned and delivered doses were compared based on target coverage and DVHs of OARs (lungs, cord, heart, and esophagus).
Target volume coverage was 97.5% in the simulated real-time delivery, as compared to 95% as planned based on AIP. Maximum cord dose, maximum esophagus dose, mean heart dose, and V20Gy of lung were comparable between the planned and delivered dose; the relative difference were 0.3%, 4.0%, 0%, and 2.8%, respectively. These results indicates AIP-based planning was a close representation of the real 4D dose delivery for both target and OARs when the breathing pattern is regular and reproducible.
A framework has been successfully established for verifying 4D dose delivery using the 4D-XCAT phantom and deformable image registration. This verification method is capable of incorporating specific patient anatomy and respiratory mechanics.
目前，由于缺乏患者解剖结构或呼吸信息，肺癌放射治疗中与呼吸运动相关的四维剂量验证技术非常有限。本研究的目的是开发一种基于XCAT体模的四维剂量验证技术，该技术整合了患者解剖结构和呼吸力学。
使用集成的Matlab程序生成具有患者特定解剖结构并允许输入呼吸参数的四维XCAT体模。从四维XCAT生成的10期四维CT、最大密度投影（MIP）和平均密度投影（AIP）图像用于设计三维适形治疗计划。通过计算每个射束在四维XCAT的每个阶段中针对固定的个体计划剂量的沉积剂量来模拟实时剂量递送。使用VelocityAl软件中实现的可变形图像配准，通过累积所有射束在所有阶段的剂量沉积来确定四维递送剂量。基于靶区覆盖情况和危及器官（肺、脊髓、心脏和食管）的剂量体积直方图（DVH）比较计划剂量和递送剂量。
在模拟的实时递送中，靶区体积覆盖率为97.5%，而基于AIP计划的覆盖率为95%。计划剂量和递送剂量之间，脊髓最大剂量、食管最大剂量、心脏平均剂量和肺的V20Gy相当；相对差异分别为0.3%、4.0%、0%和2.8%。这些结果表明，当呼吸模式规则且可重复时，基于AIP的计划是靶区和危及器官真实四维剂量递送的近似表示。
已成功建立了一个使用四维XCAT体模和可变形图像配准来验证四维剂量递送的框架。这种验证方法能够整合特定患者的解剖结构和呼吸力学。
