Magome T, Arimura H, Shioyama Y, Mizoguchi A, Tokunaga C, Nakamura K, Honda H, Ohki M, Toyofuku F, Hirata H
Kyushu University, Fukuoka, Japan.
Med Phys. 2012 Jun;39(6Part24):3908. doi: 10.1118/1.4735950.
The quality of a treatment plan for stereotactic body radiotherapy (SBRT) depends on an experience of each treatment planner. Therefore, the treatment plans are subjectively determined by comparison of several treatment plans developed by time consuming iterative manners, while considering the benefit to a tumor and the risk to the surrounding normal tissues. The aim of our study was to develop an automated optimization method for beam arrangements based on similar cases in a database including plans designed by senior experienced treatment planners.
Our proposed method consists of three steps. First, similar cases were automatically selected based on image features from the treatment planning point of view. We defined four types of image features relevant to planning target volume (PTV) location, PTV shape, lung size, and spinal cord positional features. Second, the beam angles of the similar case were registered to the objective case with respect to lung regions using a linear registration technique. Third, the beam direction of the objective case was locally optimized based on the cost function considering radiation absorption in normal tissues and organs at risk. The proposed method was evaluated with 10 test cases and a treatment planning database including 81 cases by using eight planning evaluation indices such as D95, lung V20, and maximum spinal cord dose.
The proposed method may provide usable beam directions, which have no statistically significant differences with the original beam directions (P > 0.05) in terms of the seven planning evaluation indices. Moreover, the mean value of D95 for 10 test cases was improved with a statistically significant difference by using the proposed method, compared with the original beam directions (P = 0.03).
The proposed method could be used as a computer-assisted treatment planning tool for determination of beam directions in SBRT.
立体定向体部放射治疗（SBRT）治疗计划的质量取决于每位治疗计划师的经验。因此，治疗计划是通过耗时的迭代方式制定多个治疗计划并进行比较后主观确定的，同时要考虑对肿瘤的益处和对周围正常组织的风险。我们研究的目的是基于数据库中的相似病例开发一种用于射束排列的自动优化方法，该数据库包括由经验丰富的资深治疗计划师设计的计划。
我们提出的方法包括三个步骤。首先，从治疗计划的角度基于图像特征自动选择相似病例。我们定义了与计划靶体积（PTV）位置、PTV形状、肺大小和脊髓位置特征相关的四种图像特征。其次，使用线性配准技术将相似病例的射束角度相对于肺区域配准到目标病例。第三，基于考虑正常组织和危及器官中辐射吸收的成本函数对目标病例的射束方向进行局部优化。通过使用诸如D95、肺V20和脊髓最大剂量等八个计划评估指标，对10个测试病例和一个包含81个病例的治疗计划数据库对所提出的方法进行了评估。
所提出的方法可以提供可用的射束方向，在七个计划评估指标方面与原始射束方向相比无统计学显著差异（P>0.05）。此外，与原始射束方向相比，使用所提出的方法，10个测试病例的D95平均值有统计学显著差异的改善（P = 0.03）。
所提出的方法可作为一种计算机辅助治疗计划工具，用于确定SBRT中的射束方向。
