电子射野影像装置用于监控放射治疗中初次摆位误差的研究

目的利用电子射野影像装置（EPID）对不同部位肿瘤放射治疗中的初次摆位误差进行量化分析,了解不同部位的摆位误差,了解不同医生对摆位误差的接受程度有无差异。方法 2006年1月～2007年12月接受放射治疗的患者中按头颈、胸部、腹盆腔不同部位,按不同医生治疗组随机选取58例,其中男性30例,女性28例;年龄23～86岁,平均年龄56.9岁。按部位分别为头及头颈部20例,胸部19例,腹盆部19例。采用双曝光法拍摄照射野验证片。参考图像为CT定位后的数字重建图像（DRR）或X射线模拟机定位图像,图像比较采用骨性结构为标志,勾画出参考图像上的骨性结构,与EPID进行最大程度重合后得出在水平、垂直方向上的摆位误差数据并进行分析。结果摆位误差水平方向为（1.02±1.15）mm（0~5mm）,垂直方向为（1.11±1.20）mm（0～6mm）。90%以上摆位误差小于3mm。头部或头颈部固定摆位误差最小,水平方向和垂直方向分别为（0.64±0.65）mm和（0.91±0.98）mm。不同医生组之间摆位误差无明显统计学意义。在不同部位治疗中,在垂直方向摆位误差上,头及头颈部固定明显小于胸部和腹盆部固定,差异有统计学意义（P=0.02）。结论绝大多数摆位误差在容许范围。实施治疗前,先行X射线模拟机摆位不失为一种缩小摆位误差的有效办法。

Electronic portal imaging device in the initial position error of monitoring radiation therapy

Objective To quantitatively analyze the initial position error in different positions of irradiated cancer patients by electronic portal imaging device （EPID）, and find out the acceptance in different oncologists. Methods From January 2006 to December2007,58cancerpatientswereenrolled,male30,female28,aged23-86years,meanage 56.9 years, which randomly selected from different irradiation treatment group by different positions（head and neck n = 20, chest n = 19, abdomen and pelvic n = 19 ）. Treatment planning computed tomography （CT） scans were used to create orthogonally digital reconstructed radiographs （DRR） for portal image registration, which were compared with portal images from EPID of Elekta iView GT by comparing with bone structure in horizontal and vertical directions. Results The position error of horizontal position was （1.02 ± 1.15） millimeters（0 ~ 5 millimeters）, and the vertical was （1.11 ± 1.20） millimeters（0 ~ 6 millimeters）. More than 90 % error was less than 3 millimeters. The smallest error in the horizontal direction and vertical direction were （0.64 ± 0.65） millimeter and （0.91 ± 0.98） millimeter respectively with head or head and neck fixed. No significant statistical differences were found in different radiation groups. The significant statistical difference was found in vertical position, with the head-fixed or head and neck fixed was smaller than the chest and abdominal pelvis fixed（P = 0.02）. Conclusion It is demonstrated that the majority of position error is in the permissible range. The implementation of the X-ray simulator before treatment is an effective approach to reduce position error.