摘要
目的:通过计算机编程,实现两种宫颈癌后装治疗剂量的调强优化方法,并与传统的A点优化方法进行比较。方法:选择一套已放置3根施源器(1根在宫腔,2根在隆穹)的宫颈癌患者的CT图像,采用VC++编程,读入图像并重建施源器,然后根据靶区器官的限量约束条件(处方剂量为5 Gy),自动激活部分驻点,分别采用模拟退火算法和随机最小二乘法进行调强逆向计算驻点的时间。接着,设A点(宫口上2 cm,旁开2cm)为剂量参考点,取上述相同的驻点,以均等权重方式计算驻点时间。最后比较3种优化方法的剂量分布。结果:在模拟退火法、随机二乘法和A点方法中,5 Gy覆盖靶区的体积分别为84.2%、80.2%和79.5%,靶区的D90分别为4.1 Gy、4.3 Gy和3.7 Gy,直肠D2cc为4.6 Gy、3.2 Gy和3.9 Gy,膀胱的D2cc为5.0 Gy、2.7Gy和4.1 Gy,靶区剂量的均匀度为65%、49%和45%。结论:模拟退火算法可以同时兼顾靶区受量和重要器官限量,剂量均匀性好,能满足个体化后装治疗的需求。随机最小二乘法虽能提高靶区的D90,减少直肠膀胱的受量,但相邻驻点之间的时间梯度变化较大。
Objective:To implement two intensity-modulated optimization algorithms of cervical afterloading dose with computer programming.These two algorithms would be comparied with the conventional point A optimization algorithm.Methods:Based on VC++ programming,three afterloading source applicators(ones in uterine cavity and twos in fornix) were reconstructed on a series of CT images of a patient with cervical cancer.Next,the prescribed dose was setted at 5 Gy,according to some dose constraint conditions of target and organs at risk,several dwelling points which should be used in intensity-modulated optimization process were automatically activated.Then the activation time of these above dwelling points was calculated respectively with the simulated annealing algorithm(SAA) and the random least squares algorithm(RLSA).Next,point A was setted as the dose point,the activation time of these above same dwelling ponits was calculated again with the equal weight algorithm(EWA).Finally,these dose distribution of the aboved three algorithms were comparied reciprocally.Results:In the aboved three algorithms,the coverage volume of prescription dose(5 Gy) was respectively 84.2%(SAA),80.2%(RLSA) and 79.5%(EWA),the D90 of target was respectively 4.1 Gy,4.3 Gy and 3.7 Gy,the D2 cc of rectum was respectively 4.6 Gy,3.2 Gy and 3.9 Gy,the dose homogeneity index of target was respectively 65%,49% and 45%.Conclusions:The simulated annealing algorithm could give consideration to both the absorbed dose of target regions and limited radiation dose ofvital organs,its dose distribution had good homogeneity index,so it could basically meet the individualized requirements of after loading brachy therapy.The random least squares algorithm could improved the D90 of target and reduce the radiation dose of bladder,but its time gradient was big between two adjacent dwelling points.
出处
《中国医学物理学杂志》
CSCD
2014年第5期5114-5119,共6页
Chinese Journal of Medical Physics
基金
广东省医学科研基金(B2013186)
广州市医药卫生科技项目(20131A011162)
广州医学院青年科研项目(2012A17)
关键词
宫颈癌
后装治疗
调强优化
模拟退火
最小二乘法
cervical cancer
brachytherapy
intensity-modulated optimization
simulated annealing
least squares