3种配准方式下肺癌图像引导放射治疗摆位误差的比较

目的研究滑轨CT不同配准方式对肺癌大分割调强放疗摆位误差的影响。方法选取2016年10月至2017年10月在安徽医科大学第一附属医院放疗科就诊的肺癌患者24例,每周行滑轨CT扫描所得图像与原计划图像行灰度、骨性、手动3种模式在线配准,得出X、Y、Z轴3个方向平移误差,并进行统计分析。结果灰度配准、骨性配准和手动配准方式测量的X、Y、Z轴摆位误差分别为(0. 21±0. 16)、(0. 29±0. 15)、(0. 37±0. 22) cm,(0. 27±0. 16)、(0. 35±0. 25)、(0. 29±0. 19) cm,(0. 18±0. 12)、(0. 28±0. 16)、(0. 23±0. 14) cm,结果显示Y轴平移误差最大,其次为X轴,Z轴最小,且3组数据的差异有统计学意义(P <0. 05)。其中灰度模式在X、Y、Z轴上平移误差较骨性模式均缩小,灰度模式在X、Z轴上平移误差较手动模式有所缩小,差异有统计学意义(P <0. 05)。结论肺癌患者行图像引导大分割调强放疗时,应用滑轨CT中的灰度模式可缩小摆位误差。

Comparison of IGRT set-up errors in lung cancer with three registration models

Objective To investigate the impacts of different CT-on-Rail matching models on set-up errors in hypofraction radiotherapy dealing with lung cancer. Methods Twenty-four patients were randomly selected from the First Affiliated Hospital of Anhui Medical University during Oct.2016 to Oct.2017.Each patient received CT-on-rail scan each week,and obtained images undergone online registrations with original ones in three different aspects,including grey-scale,bone and manual models.The translation deviations in X,Yand Z-axes were collected and analyzed statistically.Results The X, Y and Z axes set-up errors measured by gray-scale, bone and manual models were (0.21±0.16), (0.29±0.15), (0.37±0.22) cm respectively,(0.27±0.16), (0.35±0.25), (0.29±0.19) cm respectively, and (0.18±0.12), (0.28±0.16), (0.23±0.14)cm respectively. The results showed that the Y-axis set-up error was the largest, followed by the X-axis. The Z axis was the smallest, and the difference between three groups of data was statistically significant (P<0.05). Compared with bone model,the translation errors of grey-scale model on X,Y,Z axes were significantly reduced(P<0.05).Meanwhile,grey-scale model presented lower translation deviations on X and Z axes than manual model,which was statistically significant(P<0.05). Conclusions CT-on-rail could reduce set-up errors in lung cancer patients treated with IGIMHR,and the grey-scale model would be of top priority.