考虑标记点可见性的叶片运动轨迹算法

发布时间：2018-04-23 07:30

本文选题：放射治疗 + 跟踪　；参考：《北京协和医学院》2013年博士论文

【摘要】：背景与目的：由于人类高发肿瘤大多位于胸腹部(如肺癌),会受呼吸和胃肠蠕动等生理现象的影响,因此运动靶区的放射治疗方法是当前放疗领域的一个研究热点,目前的解决方法是4D放疗和实时跟踪放疗。本研究将两种方法有机地结合起来,在计划设计阶段确定叶片运动轨迹时,考虑标记点在将来治疗时的可见性,以便实现有计划的实时跟踪。本研究的主要研究内容为：一、在静态调强治疗模式(SMLC)下考虑标记点可见性最大化的叶片运动轨迹(即分子野序列)算法。在利用实时跟踪技术进行调强放射治疗时,我们需要在患者体内(靶区或靶区附近区域)植入标记点,以利用影像系统(如EPID)对患者靶区的运动进行实施跟踪。由于调强治疗时多叶准直器(MLC)会形成大小形状不同的照射子野,因此我们需要尽可能多的提高标记点的探测概率,即使得标记点尽可能的出现在EPID上。为了量化标记点的探测概率,本研究在国际上首次提出了标记点可见性的概念,即标记点可见时间与射野照射时间的百分比比值；二、在动态调强治疗模式(DMLC)下考虑标记点可见性最大化的叶片轨迹算法。该研究内容与“在静态调强治疗模式下考虑标记点可见性最大化的分子野序列算法”相关,只是在动态调强模式下进行叶片轨迹的优化,优化目标同样是标记点可见性的最大化。材料与方法：我们首先建立优化算法的数学模型和该算法的程序流程图,而后采用MATLAB语言编写相应的程序以实现该优化算法。为了检验所提出的优化算法的可行性、正确性和计算效率,我们采用计算机随机生成的6个测试野(大测试野20×20、中测试野10×10、小测试野5×5各2个,含有1个或3个标记点)以及临床的15个测试野(3个前列腺癌病例,每个病例均采用5野放射治疗计划设计方案,并含有3个标记点)进行算法评估。结果：对于静态调强,采用随机测试野及临床测试野进行的优化算法评估的结果表明：1)相比于初始的子野序列,优化的子野序列不会增加总实施强度级数目；2)相比于初始的子野序列,优化的子野序列提高了标记点可见性；对于动态调强,采用随机测试野及临床测试野进行的优化算法评估,也有类似的结果。另外,无论是子野序列优化算法,还是叶片轨迹优化算法,各测试野的程序运行时间均小于1s,说明本研究提出的优化算法的计算效率很高。结论：本研究在国际上首次提出了标记点可见性的概念,很好的解决了实时跟踪放射治疗过程中,量化标记点探测效率的问题。在静态调强治疗模式下,提出了考虑标记点可见性的子野序列优化算法,使得优化后子野序列的标记点可见性可以达到最大值；在动态调强治疗模式下,提出了考虑标记点可见性的叶片轨迹优化算法,使得优化后叶片轨迹的标记点可见性也可以达到最大值。本研究提出的优化算法是可行的,正确的,并且计算效率也很高,可以作为今后开展实时跟踪放射治疗的理论基础,有很高的理论价值和应用前景。
[Abstract]:Background and purpose: human high incidence tumors are mostly located in the chest and abdomen (such as lung cancer), which are affected by physiological phenomena such as respiratory and gastrointestinal peristalsis. So the radiation therapy in the target area is a hot spot in the field of radiotherapy. The current solution is 4D radiotherapy and real-time tracking radiotherapy. This study organically knot the two methods. Together, when the blade trajectory is determined at the planned design stage, the visibility of the marker points in the future treatment is considered in order to achieve a planned real-time tracking. The main research contents of this study are: the blade motion trajectory (the molecular field sequence) algorithm which considers the maximum visibility of the mark point under the static intensity modulated treatment mode (SMLC). When using real-time tracking technique for intensity modulated radiation therapy, we need to implant markers in the patient (target area or area near the target area) to use image system (such as EPID) to track the movement of the target area. As a result of the intensity modulation treatment, the multileaf collimator (MLC) will form a different size and shape of the radiation field, so we need to To improve the detection probability of the mark point as much as possible, even if the mark point is given as much as possible on the EPID. In order to quantify the detection probability of the mark point, the concept of the visibility of the mark point is first proposed in this study, that is, the percentage ratio of the visible time to the radiation time of the mark point, and two, in the dynamic intensity modulation therapy model (DMLC The blade trajectory algorithm which considers the maximum visibility of the mark point is considered. The research is related to "the molecular field sequence algorithm which considers the maximum visibility of the mark point in the static intensity modulated treatment mode", only the blade trajectory optimization is carried out in the dynamic intensity modulation mode, and the optimization target is also the maximization of the visibility of the mark point.
Materials and methods: we first set up the mathematical model of the optimization algorithm and the program flow chart of the algorithm, and then use the MATLAB language to write the corresponding program to realize the optimization algorithm. In order to test the feasibility, correctness and efficiency of the proposed optimization algorithm, we use 6 test fields randomly generated by the computer (big test). The field 20 x 20, the medium test field 10 x 10, the small test field 5 x 5 each 2, contains 1 or 3 marking points and the clinical 15 test fields (3 prostate cancer cases, each case adopts 5 radiation therapy plan design, and contains 3 marker points) to carry on the algorithm evaluation.
Results: for static strength, the results of optimization algorithms used in random test field and clinical test field show that: 1) compared to the initial subfield sequence, the optimized subfield sequence will not increase the number of total implementation intensity levels; 2) compared to the initial subfield sequence, the optimized subfield sequence improves the visibility of the marking point; The optimization algorithm of random test field and clinical test field has similar results. In addition, the running time of each field is less than 1s, whether it is the optimization algorithm of the sub field sequence or the optimization algorithm of the blade path, which shows that the efficiency of the optimization algorithm proposed in this study is very high.
Conclusion: the concept of mark point visibility is first proposed in this study, which solves the problem of the detection efficiency of quantized marking points in the process of real-time tracking radiation therapy. In the static intensity modulated treatment mode, a subfield sequence optimization algorithm, which considers the visibility of the mark point, is proposed, which makes the marker points of the optimized subfield sequence visible. In the dynamic intensity adjustment treatment model, the blade trajectory optimization algorithm considering the visibility of the mark point is proposed, which makes the mark point visibility of the optimized blade trajectory can reach the maximum. The optimization algorithm proposed in this study is feasible, correct and highly efficient, which can be carried out in the future. The theoretical basis of real-time tracking radiation therapy has high theoretical value and application prospect.

【学位授予单位】：北京协和医学院
【学位级别】：博士
【学位授予年份】：2013
【分类号】：R730.55

【相似文献】