基于形变配准的头颈部肿瘤自适应IMRT方案的剂量学研究

发布时间：2018-06-30 01:50

本文选题：基于梯度的形变配准 + IMRT　；参考：《山东师范大学》2017年硕士论文

【摘要】：放射治疗是目前最为流行且有效的肿瘤治疗方法,由于IMRT具有陡峭的剂量跌落梯度,精准的剂量调节,更好的危及器官保护能力,所以广泛应用在头颈部肿瘤放射治疗当中。临床上,癌症患者的治疗周期一般为6-7周,在此期间,癌症病人的体重变化,以及肿瘤体积的变化等因素很可能导致靶区、OARs的解剖结构发生变化。如果继续使用原始计划做放射治疗,就会使靶区剂量降低,危及器官剂量升高,进而促使正常组织受到射线的损伤,增加并发症发生概率,大大降低病人的生活质量。基于以上情况,目前临床上采取的措施是在肿瘤患者每次放射治疗前,进行CT或者CBCT扫描来指导摆位,治疗3-4周时修改一次治疗计划。然而仅仅修改一次计划还不能很好的弥补癌症患者靶区、OARs的解剖结构变化对剂量分布造成的严重影响。因此本论文针对目前临床上存在的上述疑难问题,设计四种自适应IMRT调整方案,巧妙运用基于梯度的形变配准方法,解决临床放疗中的难题,主要内容概括为以下两个方面:(1)基于梯度的形变配准方法的研究本论文针对CBCT图像密度分辨率低,以及CBCT图像与CT图像密度不一致的问题,提出了一种基于梯度的形变配准方法。本部分首先对CT与CBCT图像的成像原理,成像特点展开了论述,并针对以上两种图像的特点,构建空间形变模型,设计高效的形变配准方法,并运用临床病人数据对该方法进行了实验验证分析,为了进一步证明该形变配准方法的有效性及可行性,本论文将基于梯度的配准算法与配准领域内公认的“金标准”Demons形变配准算法进行了对比试验、分析。(2)四种头颈部肿瘤自适应IMRT调整方案的对比研究本论文,首先选取28例头颈部肿瘤患者进行试验研究,所有肿瘤患者均接受常规分割治疗。在放疗周期内,选取患者各个放疗周第1天扫描获取的CBCT图像(共6周)作为参考图像,计划CT图像作为目标图像,将两者形变配准后的图像定义为“伪CT”图像。其次,设计四种不同频次计划调整方案,无计划调整作为参考方案,定义每种方案的实际剂量与计划调整剂量,并在所得到的6套“伪CT图像”上重新计算剂量,并运用基于密度的形变配准算法进行单周分次剂量的叠加,计算得到每种计划调整方案的累加实际剂量与计划调整剂量。最后,量化评估放疗周期期间肿瘤靶区、OARs体积与剂量的变化,并分别统计分析两者的变化规律,以及两者变化之间的相关性。最终得出结论:再计划可确保靶区足够剂量,使危及器官受量控制在安全范围内,再计划频次越高,实际剂量越接近计划调整剂量,2次再计划方案效率最高。
[Abstract]:Radiation therapy is the most popular and effective method of cancer treatment. Because IMRT has a steep dose drop gradient, accurate dose regulation, and better endanger the organ protection, it is widely used in head and neck cancer radiotherapy. In clinical, cancer patients are usually treated for 6-7 weeks. During this period, cancer patients The changes in body weight and tumor volume may lead to the changes in the anatomical structure of the OARs. If the original plan is used for radiation therapy, the dose of the target area will be reduced and the dose of the organ is increased. Then the normal tissue is damaged by radiation, the probability of complication is increased, and the patient is greatly reduced. Quality of life. Based on the above situation, the current clinical measures are to conduct CT or CBCT scans before each radiotherapy for the cancer patients to guide the pendulum, and to modify a treatment plan at 3-4 weeks. However, only a revision of the plan can not make up for the target area of the cancer patients, and the changes of the anatomical structure of OARs to the dose distribution In this paper, four adaptive IMRT adjustment schemes are designed to solve the difficult problems existing in the clinic and apply the gradient based deformation registration method to solve the problems in clinical radiotherapy. The main contents are as follows: (1) research on the paper based on the gradient based deformation registration method. For the low density resolution of CBCT image and the disagreement between the CBCT image and the density of the CT image, a gradient based deformation registration method is proposed. This part first discusses the imaging principle of CT and CBCT images and the imaging characteristics, and constructs a spatial deformation model for the characteristics of the above two images, and designs a highly effective deformation registration. In order to further prove the validity and feasibility of the deformation registration method, this paper makes a comparative experiment based on the gradient registration algorithm and the recognized "gold standard" Demons deformation registration algorithm in the field of registration. (2) four kinds of head and neck tumors. A comparative study of the adaptive IMRT adjustment scheme, 28 cases of head and neck tumors were selected and all the patients received routine segmentation. During the radiotherapy period, the CBCT images (6 weeks) obtained by the first day scans of each radiotherapy week were selected as the reference images, and the CT images were planned as the target images, and two were planned. The image after deformation registration is defined as "pseudo CT" image. Secondly, four different frequency planning schemes are designed, without planned adjustment as a reference scheme, the actual dose and planned dose of each scheme are defined, and the dosage is recalculated on the 6 set of "pseudo" images, and the density based deformation registration algorithm is used. On the basis of the superposition of the single weekly dose, the cumulative actual dose and the planned adjusted dose of each plan adjustment scheme were calculated. Finally, the changes of the tumor target area, OARs volume and dose during the radiotherapy period were quantified, and the changes of the two changes were statistically analyzed, and the correlation between the two changes were analyzed. Finally, the conclusion was drawn. It can ensure that the target area is sufficient enough to control the organ tolerance within the safe range. The more the planned frequency is higher, the more the actual dose is closer to the planned adjustment dose, the 2 re plan is the most efficient.
【学位授予单位】：山东师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R739.91;TP391.41

【参考文献】