Effect of deformable registration on the dose calculated in radiation therapy planning CT scans of lung cancer patients
Effect of deformable registration on the dose calculated in radiation therapy planning CT scans of lung cancer patients
复制标题
DOI：
10.1118/1.4903267
复制
发表时间：
2015-01-01
期刊：
影响因子：
3.8
通讯作者：
Al-Hallaq, Hania A.
中科院分区：
文献类型：
--
作者：
Cunliffe, Alexandra R.;Contee, Clay;Al-Hallaq, Hania A.
目的：研究系列计算机断层扫描(CT)可变形图像配准对治疗计划扫描计算的辐射剂量的影响。方法：对18例接受根治性剂量(~gt；=60Gy，2Gy/次)光子放射治疗的肺癌患者进行回顾性分析。对于每个患者，收集了一次诊断性的治疗前(4-75天)CT扫描和一次带有相关剂量图的治疗计划扫描。为了建立扫描对之间的对应关系，研究人员手动确定了两次扫描之间在解剖学上对应的标志点对。然后，使用恶魔可变形配准算法和Fraunhofer MEVIS算法的两个变体(“Fast”和“Emree10”)，将治疗前扫描与计划扫描(以及相关的剂量图)共同配准。使用从三种算法中的每一种算法输出的位移矢量场，将每个治疗前扫描中的地标点自动映射到计划扫描。计算人工标记点与自动标记点之间的欧几里得距离(d(E))和计划剂量的绝对差值(垂直条Delta D垂直条)。利用回归建模，将垂直条形Delta D垂直条形建模为8像素邻域内的d(E)、剂量(D)、剂量标准差(SDudose)的函数，并使用配准算法。结果：确定了1400多个标志点对，每个患者确定了58-93(中位数：84)个点。横跨患者的平均垂直条Delta D垂直条为3.5Gy0.9~10.6Gy.使用Fraunhofer MEVIS EMPIRE10算法的注册准确率最高，患者的平均d(E)为5.2 mm(而其他两种算法的d(E)为7 mm)。因此，使用Fraunhofer MEVIS帝国10算法，平均垂直条形图Delta D垂直条形图也是最低的。垂直条形增量D垂直条形显著增加，作为d(E)(0.42GY/mm)、D(0.05GY/GY)、SD剂量(1.4GY/GY)的函数，并且所使用的算法
Purpose: To characterize the effects of deformable image registration of serial computed tomography (CT) scans on the radiation dose calculated from a treatment planning scan. Methods: Eighteen patients who received curative doses (>= 60 Gy, 2 Gy/fraction) of photon radiation therapy for lung cancer treatment were retrospectively identified. For each patient, a diagnosticquality pretherapy (4-75 days) CT scan and a treatment planning scan with an associated dose map were collected. To establish correspondence between scan pairs, a researcher manually identified anatomically corresponding landmark point pairs between the two scans. Pretherapy scans then were coregistered with planning scans (and associated dose maps) using the demons deformable registration algorithm and two variants of the Fraunhofer MEVIS algorithm ("Fast" and "EMPIRE 10"). Landmark points in each pretherapy scan were automatically mapped to the planning scan using the displacement vector field output from each of the three algorithms. The Euclidean distance between manually and automatically mapped landmark points (d(E)) and the absolute difference in planned dose (vertical bar Delta D vertical bar) were calculated. Using regression modeling, vertical bar Delta D vertical bar was modeled as a function of d(E), dose (D), dose standard deviation (SDdose) in an eight-pixel neighborhood, and the registration algorithm used. Results: Over 1400 landmark point pairs were identified, with 58-93 (median: 84) points identified per patient. Average vertical bar Delta D vertical bar across patients was 3.5 Gy (range: 0.9-10.6 Gy). Registration accuracy was highest using the Fraunhofer MEVIS EMPIRE10 algorithm, with an average d(E) across patients of 5.2 mm (compared with >7 mm for the other two algorithms). Consequently, average vertical bar Delta D vertical bar was also lowest using the Fraunhofer MEVIS EMPIRE 10 algorithm. vertical bar Delta D vertical bar increased significantly as a function of d(E) (0.42 Gy/mm), D (0.05 Gy/Gy), SDdose (1.4 Gy/Gy), and the algorithm used