不同重建方式下CT层厚对胸部肿瘤体积和剂量的影响

摘要

目的:基于两种不同的重建方式,分析CT层厚对胸部肿瘤靶区体积、剂量和危及器官受量的影响.方法:随机选取11例胸部肿瘤患者(主要为食管癌、肺癌和乳腺癌),获取层厚2 mm的CT影像(2-CT)并将其传至放疗计划系统(Eclipse 10.0)中,分别基于靶区最大层面(方法1)和定标金属小球所在层面(方法2)作为起始层进行重建,得到层厚为4、6、8、10 mm的CT影像(4-CT、6-CT、8-CT、10-CT).由医生在2-CT影像上勾画靶区和危及器官,并将其复制到其余层厚的影像中.同样基于2-CT影像中设计调强放疗(IMRT)计划,随后将其复制到其余层厚影像中重新计算.统计两种重建方法得到的各层厚影像中的靶体积和剂量,并以2-CT数据为基准加以比较.结果:两种重建方式中,靶体积的相对变化随着层厚的增大逐渐增大.当靶体积为1.3 cm3时,10-CT中相对偏差两种方法均高达84.62%,8-CT时两种方法相对偏差分别为38.46%、84.62%;体积为30~100 cm3时,各层厚的体积测量偏差方法1均小于方法2;体积>100 cm3时,两种方法中靶体积基本不随层厚而改变.适形指数相对变化随着层厚的变大而变大,对于小体积靶区(＜30 cm3),两种方法在10-CT中相对变化率分别为13.60%、11.18%,在8-CT时为6.56%、13.18%,在6-CT和4-CT时均小于等于5.08%;体积在30~100 cm3时,方法1相对变化率为1.29%~3.03%,方法2为1.34%~5.42%;大体积靶区受层厚的影响小,均＜1.38%.两种重建方法的均匀性指数相对变化率只在小体积时随层厚增大有小幅度增加,在8-CT、10-CT时相对偏差较大,方法1为12.95%、17.42%,方法2为16.15%、15.43%.两种重建方式得到的IMRT计划中,脊髓的最小值、平均值,肺的V5、V30,心脏的V30、V40受层厚影响显著(P＜0.05).结论:在胸部肿瘤放疗中,基于靶区最大层面为起始层的重建方式优于基于定标金属小球所在层面为起始层的重建.同时,本研究可以为改进现有的CT扫描流程提供依据.%Objective To evaluate the effects of different computed tomography (CT) slice thicknesses on the volume and dose of target areas and organs-at-risk of patients with chest cancer based on two different reconstruction methods.Methods Eleven patients with chest cancer(mainly,esophagus cancer,lung cancer and breast cancer)were enrolled in the study.The CT images with 2 mm slice thickness(2-CT)were transferred to Eclipse 10.0 treatment planning system.The images with different slice thicknesses(4,6,8 and 10 mm),namely 4-CT,6-CT,8-CT and 10-CT,were reconstructed by two methods.In Method 1,the largest target layer was taken as the starting layer,while in Method 2,the metal marker layer was treated as the starting layer.The target areas and organs-at-risk were delineated on 2-CT images and then were copied slice by slice on CT images with different slice thicknesses.Similarly,intensity-modulated radiotherapy(IMRT)plans were developed on the 2-CT images and then were copied on the other CT images.The target volume and doses were computed independently for each image. The results on 2-CT images were chosen as the reference standard. Results The target volume in two reconstruction methods was gradually increasing with the increase of slice thickness.When the target area was 1.3 cm3,the volume deviation on 10-CT images in two methods was up to 84.62% and the deviation on 8-CT images was 38.46% (Method 1) and 84.62% (Method 2). For the target volume of 30-100 cm3, the volume changes in Method 1 were significantly smaller than that in Method 2;for target volume larger than 100 cm3,the deviation was trivial,without being increasing with the increase of slice thickness.The changes in radiation conformal index(CI)were gradually larger as the thickness increased.The changes in CI of small targets(＜30 cm3)were particularly significant on 8-CT image(6.56% and 13.18% for Method 1 and Method 2, respectively) and 10-CT image (13.60% and 11.18% for Method 1 and Method 2, respectively)and that for larger targets with 30-100 cm3were 1.29%-3.03%(Method 1)and 1.34%-5.42%(Method 2).Both the variation tendency of CI in Method 1 and Method 2 was less than 1.38%.Nevertheless,the difference of HI between two methods was not obvious. The changes of HI were larger on 8-CT image (12.95% and 16.15% for Method 1 and Method 2,respectively)and 10-CT image(17.42% and 15.43% for Method 1 and Method 2,respectively).The minimum dose,average dose of spinal cord,the V5,V30of lung,and the V30,V40of heart in IMRT were significantly affected by CT slice thickness (P＜0.05). Conclusion In the IMRT for chest cancer, the reconstruction method based on the largest target layer is superior to that based on the metal marker layer.Meanwhile,the study provides some reference for improving the existing CT scanning process.