A Limited-angle Intra-fractional Verification (LIVE) System for SBRT Treatments

描述（由申请人提供）： 立体定向放射治疗（SBRT）已成为早期非小细胞肺癌（NSCLC）患者的有效治疗模式。肺SBRT的成功取决于靶定位的准确性以及关键正常结构的保留。 由于肺SBRT的PTV边缘较窄、分数dse较高且分数较少、治疗时间较长，因此分次内验证对于肺SBRT至关重要。Purdie等人表明，肺肿瘤分次内运动随治疗时间增加，SBRT可超过1 cm。Zhao等人报告，由于分次内运动，使用5 mm边缘的肺SBRT中PTV覆盖率从95%最大下降至78%。使用kV荧光透视和MV电影成像进行分次内验证的准确性有限，尤其是对于小肿瘤，因为2个图像中的结构重叠。常规CBCT由于扫描角度大、采集时间长、成像剂量高和机械间隙有限，在分次内验证方面的应用也有限。到目前为止，3D或4D体积X射线成像技术都不可用于在实际治疗输送期间进行目标验证。本研究计划提出开发一种有限角度分次内验证（LIVE）系统，用于在弧形治疗输送期间或在3D/IMRT射束之间进行快速4D验证。LIVE系统的长期目标是通过大幅减少分次内运动引起的治疗误差和提高ART的准确性来改善肺SBRT的肿瘤控制和减少正常组织毒性。该提案的具体目标是：1）通过开发新的图像采集，分选，kV-MV聚合和重建技术来构建LIVE系统。2)通过呼吸运动建模、新颖的加速策略以及XCAT和Monte Carlo（MC）模拟，优化LIVE的准确性、效率和成像剂量。3)评估LIVE系统用于IGRT和ART。LIVE创造性地结合了许多正在开发的用于快速分次内验证的新技术：1）有限角度kV-MV图像采集; 2）基于投影傅立叶变换的自动投影排序; 3）使用线性拟合、虚拟单能量和双CT技术的kV-MV聚合; 4）4D和相位匹配的数字断层合成（TOM）和基于先验知识的CBCT重建方法; 5）呼吸运动建模; 6）自适应重建策略;以及7）新的加速策略。将通过4D数字扩展心脏躯干体模（XCAT）和蒙特卡罗（MC）模拟针对不同肿瘤尺寸和位置优化LIVE系统的参数，包括射束能量、注量、kV-MV成像序列、kV-MV投影数量和扫描角度。MC还将用于评价成像剂量。将通过使用CIRS动态胸部运动体模的体模研究和使用回顾性研究中收集的肺部患者CBCT和MV电影图像数据的患者研究，定量评价LIVE系统用于IGRT和ART应用的有效性。

DESCRIPTION (provided by applicant): Stereotactic body radiation therapy (SBRT) has become an effective treatment paradigm for early stage non- small cell lung cancer (NSCLC) patients. The success of lung SBRT is determined by the accuracy of target localization as well as sparing of critical normal structures. Intrafraction verification is critical for lung SBRT due to its tight PTV margin, high fractional dse with small fraction number, and long treatment time. Purdie et al showed that lung tumor intrafraction motion increases with treatment time, and can be over 1cm for SBRT. Zhao et al reported a maximal drop of PTV coverage from 95% to 78% in lung SBRT using 5mm margin due to intrafraction motion. The accuracy of using kV fluoroscopy and MV cine imaging for intrafraction verification is limited especially for small tumors due to overlaying structures in 2 images. Conventional CBCT also has limited application for intrafraction verification due to large scanning angle, long acquisition time, high imaging dose and limited mechanical clearance. By far, neither 3D nor 4D volumetric x-ray imaging techniques are available for target verification on-the-fly during actual treatment delivery. This research plan proposes to develop a Limited-angle Intrafraction Verification (LIVE) system for fast 4D verification during arc treatment delivery or in between 3D/IMRT beams. The long term goal of the LIVE system is to improve the tumor control and reduce normal tissue toxicity for lung SBRT by substantially reducing the treatment errors caused by intrafraction motion and improving the accuracy of ART. The specific aims of the proposal are: 1) Build the LIVE system by developing novel image acquisition, sorting, kV-MV aggregation and reconstruction techniques. 2) Optimize the accuracy, efficiency and imaging dose of LIVE through respiratory motion modeling, novel acceleration strategies, and XCAT and Monte Carlo (MC) simulation. 3) Evaluate the LIVE system for IGRT and ART. LIVE creatively combined a number of new technologies being developed for fast intrafraction verification: 1) limited-angle kV-MV image acquisition; 2) automatic projection sorting based on Fourier Transformation of projections; 3) kV-MV aggregation using linear fitting, virtual mono-energetic and dual CT techniques; 4) 4D- and phase-matched digital tomosynthesis (DTS) and prior knowledge based CBCT reconstruction methods; 5) respiratory motion modeling; 6) adaptive reconstruction strategy; and 7) new acceleration strategies. Parameters of the LIVE system, including beam energy, fluence, kV-MV imaging sequence, number of kV-MV projections, and scan angle will be optimized for different tumor sizes and locations through 4D digital extended cardiac-torso phantom (XCAT) and Monte Carlo (MC) simulation. MC will also be used to evaluate the imaging dose. The effectiveness of LIVE system for IGRT and ART applications will be quantitatively evaluated through phantom studies using the CIRS dynamic thorax motion phantom and patient studies using lung patient CBCT and MV cine image data collected in a retrospective study.