论文部分内容阅读
目的探讨头颈部肿瘤锥形束CT图像不同配准方式对摆位误差的影响。方法选取2010年6月至2012年6月应用锥形束CT图像的头颈部肿瘤患者74例,随机分为观察组和对照组,每组各37例。对照组采取手动微调配准方式,观察组采取自动灰度结合手动微调配准方式,比较两组平移误差和旋转误差。结果观察组头脚方向误差为(-0.02±0.15)mm,左右方向误差为(0.06±0.37)mm,前后方向误差为(-0.02±0.21)mm;对照组头脚方向误差为(-0.05±0.31)mm,左右方向误差为(0.18±0.62)mm,前后方向误差为(-0.04±0.36)mm。观察组平移误差显著低于对照组,差异有统计学意义(P<0.05);观察组头脚方向误差为(0.41±1.26)°,左右方向误差为(0.13±1.15)°,前后方向误差为(0.14±0.92)°;对照组头脚方向误差为(0.75±2.63)°,左右方向误差为(0.24±1.38)°,前后方向误差为(0.26±1.07)°。观察组旋转误差显著低于对照组,差异有统计学意义(P<0.05)。结论锥形束CT图像采取自动灰度结合手动微调配准方式,可以减少摆位误差,提高治疗精确性、保证治疗质量。
Objective To investigate the effect of different registration methods of cone beam CT images on the setup error of head and neck tumors. Methods From June 2010 to June 2012, 74 patients with head and neck cancer using cone beam computed tomography were randomly divided into observation group and control group, 37 cases in each group. In the control group, the manual fine-tuning registration method was adopted. In the observation group, automatic grayscale combined with manual fine-tuning registration method was used to compare the translation errors and rotation errors of the two groups. Results The head-to-tail error was (-0.02 ± 0.15) mm in the observation group, (0.06 ± 0.37) mm in the left-right direction, and (-0.02 ± 0.21) mm in the anterior-posterior direction. The head- 0.31) mm, the error in left and right direction was (0.18 ± 0.62) mm, and the error in front and back direction was (-0.04 ± 0.36) mm. The translation error of the observation group was significantly lower than that of the control group (P <0.05). The error of head and foot direction was (0.41 ± 1.26) ° in the observation group and (0.13 ± 1.15) ° in the left and right direction, (0.14 ± 0.92) °. The head-to-tail direction error of the control group was (0.75 ± 2.63) °, the left-right direction error was (0.24 ± 1.38) ° and the anteroposterior direction error was (0.26 ± 1.07) °. The rotation error of the observation group was significantly lower than that of the control group, the difference was statistically significant (P <0.05). Conclusion Conical beam CT images with automatic grayscale combined with manual fine-tuning registration can reduce the positioning error, improve the accuracy of treatment, to ensure the quality of treatment.