基于断层调强放射治疗中扇形束兆伏级CT数据分析鼻咽癌的摆位误差

An analysis on set-up errors by data of mugavoltage fan-beam computed tomography during intensity-modulated radiotherapy for nasopharyngeal carcinoma

目的:基于断层调强放射治疗中扇形束兆伏级断层扫描数据分析鼻咽癌治疗分次间的摆位误差,探讨影响摆位误差的因素。方法:37例鼻咽癌患者于2015年2-9月期间接受断层调强放射治疗。所有患者分别于每日常规摆位后、断层调强放射治疗前行断层扫描,并与计划断层扫描匹配,获取误差数据。通过MPTV=2.5∑+0.7σ(MPTV:计划靶体积边界;∑:系统误差;σ:随机误差)计算MPTV。结果:患者校正前三维方向和旋转角度的误差绝对值均值分别为左右(2.102±0.040 6)mm;头尾方向(1.490±0.034 8)mm;前后方向(1.306±0.335)mm和旋转(1.392±0.038 4)度。对应的各方向MPTV为左右方向为3.467 5 mm,头尾方向为2.979 5 mm,前后方向为2.888 5 mm。分次间三维方向位移随时间逐渐增加且差异有统计学意义(P<0.05)。单因素分析显示体重下降和颈部淋巴结退缩影响摆位误差。结论:鼻咽癌调强放疗计划设计时均匀外放3 mm得到PTV并不合适,应个体化设计;同时摆位误差随治疗时间逐渐增大,在线纠正摆位误差保证了断层调强放疗的精确实施。

Objective To explore the inter-fraction setup errors and affecting factors from data of daily fan- beam megavohage computed tomography （MVCT）. Methods A total of 37consecutive NPC patients treated with tomotherapy were hospitalized during the period of February 2015 to September 2015. For each patient, one MVCT scan was obtained after conventional positioning, online correction and tomotherapy delivery daily, and the scans were put into the planning computed tomography to determine inter-fraction setup errors. The MPTV was calculated with the equation： MPTV = 2.5∑± 0.7σ（ ∑ ： systematic error; σ： random error）. Results The average absolute errors of the inter-fraction were （2.102 ±0.040 6） mm, （1.490 ± 0.034 8） mm, （1.306 ±0.335） mm and （1.392 ± 0.038 4）° at three dimensions. The total MPTV accounting for inter-error was 3.467 5 mm, 2.979 5 mm, and 2.888 5 mm. Gradual increases in both inter-fraction three-dimensional displacement were observed with time and treatment （P 〈 0.05 ）. Univariate analysis revealed that weight loss and retraction of neck lymph nodes were affecting factors of set-up errors. Conclusions 3 mm margins uniformly expended from clinical target volume to planning target volume may not be suitable. The personalized margin should be adopted for the design of IMRT planning. Displacement increases as a treatment course is prolonged.