IMRT逆向计划中优化算法的研究

发布时间：2018-05-22 17:57

本文选题：IMRT + 逆向计划　；参考：《东北大学》2009年硕士论文

【摘要】：IMRT是一种先进的三维放射治疗技术,近年来广泛用于临床肿瘤治疗。它的特点在于能实现几何形态和强度分布的双重适形,从而提高治疗比,降低对病人的伤害。与传统的正向计划不同,IMRT采用逆向计划,先由医生确定处方剂量,然后由计算机给出实现该结果的方法和参数,从而实现了治疗计划的自动最佳优化。本文根据IMRT逆向计划的研究现状,结合现代最优化方法,对IMRT逆向计划的优化作了较为深入的研究。首先介绍了规则野、不规则野的各种剂量计算模型,其中着重论述了目前常用的基于笔形束模型的剂量计算方法,用于计算由多叶准直器形成的不规则野的剂量分布；然后研究了IMRT逆向计划优化算法,针对模拟退火算法和单纯形法的特点,将二者相结合,形成了一种改进的模拟退火算法：SMSA算法,并用该算法实现了子野形状和权重的优化。实验结果表明,该算法不仅能很好的收敛于最优值,而且相比于模拟退火算法,优化速度大大提高。同时,优化过程中,为更好的保护靶区周围的正常组织及减少计算量,构建了新的优化模型；在剂量均匀性约束和组织重要程度约束的条件下,建立了新的目标函数。最后,进行了实验测试,并用剂量一体积直方图和等剂量线进行了计划评估。临床模拟病例结果表明,采用本文的算法及建立的目标函数,达到了临床要求,不仅提高了计划靶区的剂量均匀性,减少了正常组织的受照剂量,而且大大减少了逆向计划优化时间。本文对算法及目标函数的改进具有较好的实际意义和临床应用价值。
[Abstract]:IMRT is an advanced three-dimensional radiotherapy technique, which has been widely used in clinical tumor therapy in recent years. It is characterized by its dual conformal geometry and intensity distribution, which can improve the treatment ratio and reduce the injury to patients. Different from the traditional forward plan, IMRT adopts reverse plan. First, the prescription dose is determined by the doctor, then the method and parameters to realize the result are given by the computer, thus the automatic optimal optimization of the treatment plan is realized. According to the research status of IMRT reverse plan and modern optimization method, this paper makes a deep research on the optimization of IMRT reverse plan. In this paper, the dose calculation models of regular field and irregular field are introduced firstly, and the dose calculation methods based on pen beam model are emphatically discussed, which are used to calculate the dose distribution of irregular field formed by multi-leaf collimator. Based on the characteristics of simulated annealing algorithm and simplex method, an improved simulated annealing algorithm, called "IMRT" algorithm, is proposed to optimize the shape and weight of the subfield. Experimental results show that the proposed algorithm not only converges to the optimal value, but also improves the optimization speed compared with the simulated annealing algorithm. At the same time, in order to better protect the normal tissue around the target area and reduce the amount of calculation, a new optimization model is constructed, and a new objective function is established under the condition of dose uniformity constraint and tissue importance constraint. Finally, the experiment was carried out, and the plan was evaluated with dose-volume histogram and equal dose line. The results of clinical simulation show that the proposed algorithm and the established objective function can not only improve the dose uniformity of the planned target area, but also reduce the radiation dose of normal tissue. Moreover, the optimization time of reverse planning is greatly reduced. This paper has good practical significance and clinical application value for the improvement of algorithm and objective function.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2009
【分类号】：TP301.6;R311;O224

【参考文献】