基于靶区与危及器官重叠度的自动治疗计划设计

Automatic treatment planning based on the degree of overlap between OAR and PTV

目的:根据前列腺癌患者的解剖结构建立微分剂量体积直方图(DVH)数学模型来预测危及器官(OAR)子块的DVH,通过精准预测OAR的DVH并自动设计计划来提高治疗计划的质量和效率。方法:从17例前列腺癌计划中随机选择9例作为训练集,其余为测试集。将靶区外扩到覆盖OAR范围并将外扩区域中的OAR依次分隔成层厚3 mm的子块,根据偏正态高斯函数建立子块微分DVH的精确模型,分析靶区和OAR的重叠度与子块微分DVH关系。利用建立在Pinnacle3上的脚本程序获取测试集8例患者OAR各子块,并通过预测各OAR微分DVH来指导计划。结果:靶区与OAR重叠度越小,微分DVH的变化规律越明显。与原手工计划相比,自动计划的均匀性指数和适形度指数稍优于原手工计划(t=-1.871、3.742;P<0.05)。自动计划靶区的V_(100%)降低,V_(95%)提高,差异均无统计学意义(P>0.05)。自动计划的膀胱V_(70)、V_(60)、V_(50)(t=-2.471、-3.439、-2.376;P<0.05)以及V_(40)(P>0.05)降低。自动计划的直肠V_(70)、V_(60)、V_(50)以及_(V40)(t=-2.540、-3.416、-2.666、-2.777;P<0.05)降低。相关性分析中,自动计划与原手工计划的膀胱V_(70)、V_(60)、V_(50)和V40的差异值与重叠度呈负或正相关(相关系数=-0.357、-0.976、-0.857、0.381;P=0.385、0.000、0.007、0.352),V_(70)和V40的差异值与重叠度无统计学意义。结论:基于靶区与OAR重叠度的精确DVH预测模型的前列腺癌自动治疗计划设计使得靶区剂量提高,OAR照射剂量降低。

Objective To establish a mathematical model of differential dose-volume histogram(DVH)according to the anatomical structure of prostate cancer patients for predicting the DVH of organs-at-risk(OAR)sub-blocks,and to improve the quality and efficiency of treatment plan by the accurate prediction of the DVH of OAR and automatic treatment planning.Methods Nine of 17 prostate cancer patients were randomly selected as training set and the rest as test set.Planning target volume(PTV)was expanded to cover OAR,and the OAR in the expanded area was successively separated into sub-blocks with a thickness of 3 mm.According to the skewed normal Gaussian function,a precise mathematical model of sub-block differential DVH was established for analyzing the relationship between the degree of overlap between OAR and PTV and sub-block differential DVH.The script program based on Pinnacle3 was used to obtain the OAR sub-blocks of 8 patients in test set,and each OAR differential DVH was predicted to guide the plan.Results The smaller the degree of overlap between OAR and PTV was,the more obvious the variation of differential DVH was.The homogeneity index and conformity index of automatic plan were superior to those of original plan(t=-1.871,3.742;P<0.05).The V_(100%)of PTV in automatic plan was decreased,while V_(95%)was increased,but the differences were not statistically significant(P>0.05).Compared with those in original plan,the V_(70),V_(60),V_(50)(t=-2.471,-3.439,-2.376;P<0.05)and V_(40)(P>0.05)of the bladder in automatic plan were lower;and the V_(70),V_(60),V_(50) and V_(40)(t=-2.540,-3.416,-2.666,-2.777;P<0.05)of the rectum were also decreased.The correlation analysis showed that the differences of the V_(70),V_(60),V_(50) and V_(40) of the bladder between automatic plan and original plan were negatively or positively correlated with the degree of overlap(correlation coefficient=-0.357,-0.976,-0.857,0.381;P=0.385,0.000,0.007,0.352),but there was no statistical significance between the differences of V_(70) and V_(40) and the degree of overlap.Conclusion The automatic treatment plan for prostate cancer designed using the precise DVH prediction model based on the degree of overlap between OAR and PTV increases target dose and decreases OAR irradiation dose.