基于形变配准技术对放疗前后靶区变化的评估

发布时间：2018-01-11 07:14

本文关键词：基于形变配准技术对放疗前后靶区变化的评估　出处：《山东师范大学》2017年硕士论文　论文类型：学位论文

更多相关文章： ~(18)F-FDG PET-CT 肺癌 重叠率 形变配准

【摘要】：在根治性放化疗后,患有II或III期非小细胞肺癌(NSCLC)的患者的生存概率仍然较低。在放疗肿瘤目标体积内仍存在许多复发区域,因而表明放射治疗中的总处方剂量不足。如果加大剂量必须对照射靶区有充分的认识,即要有组织信息,还要有新陈代谢信息。然而多模态的图像PET/CT就包含了新陈代谢信息和组织信息,期望通过PET/CT图像来指导放射治疗计划。PET图像还有个独特的性质,就是可以对肿瘤区域利用标准吸收值(Standard Uptake Value,SUV)值(90%、80%、70%、60%、50%、34%SUVmax)进行勾画不同的区域,如果在放疗过程中找到对射线不敏感的SUV值区域,在制定治疗计划时就可以考虑在此区域加大剂量。患者在RT期间,大体积的肿瘤收缩或变形,体重减轻或扫描期间身体位置的变化,都有可能导致重大解剖结构的形变。形变配准是医学图像处理的常用方法,配准技术就是有目的将两幅图查找一种变换,变换是为了让两幅图的差别达到极小的程度,可以进一步应用于器官和靶区轮廓的形变以及剂量分布的计算。本论文研究就采用Symmetric Demons多分辨率形变配准的方法提取放疗前后CT的图像的作用域,并将其作用于放疗前PET图像,用来形变放疗前PET图像,使放疗前和放疗后的PET图像融合更加精准。与放疗后的PET图像相比较,通过设定图像中SUV(standard uptake value)的阈值,判断勾画轮廓之间的重叠率,以获得图像中的高摄取区域,回顾性指导精确放疗。本研究对22例非小细胞肺癌患者的PET/CT数据进行试验分析、评估,得出结论:基于氟代脱氧葡萄糖(Fludeoxyglucose,FDG)高摄取区域的高重叠率结果,表明病变区域在放疗前后的位置相对稳定。经过千伏级或兆伏级X射线照射(即RT)后的残余肿瘤基本上位于放疗前靶区对~(18)F-FDG葡萄糖的高吸取区域。初步实验结果证明本研究可用于判断靶区区域对放疗的反应,回顾性指导放疗计划,可以针对放疗后残余的靶区加大照射剂量,保护危及器官和组织,精确放疗。
[Abstract]:After radical radiotherapy and chemotherapy, the survival probability of NSCLC patients with stage II or III NSCLC was still low. This indicates that the total prescribed dose in radiotherapy is insufficient. If the dose is increased, the target area of irradiation must be fully understood, that is, there should be organized information. There is also metabolic information. However, the multimodal image PET/CT contains both metabolic and organizational information. It is expected that PET/CT images can be used to guide radiotherapy planning. Pet images also have a unique property. In other words, we can use standard Uptake value (SUV) for the tumor area. (90%) and 80% (70%) and 60% (50%). 34 / SUVmax. draw different areas if you find a radiosensitive SUV value area during radiation therapy. Increased doses in this area may be considered when planning treatment. During RT, patients with large tumors contract or deform, weight loss or changes in body position during scanning. Deformation registration is a common method of medical image processing, and registration technology is to find a transformation between two images purposefully. The purpose of the transformation is to make the difference between the two pictures to a minimum. It can be further applied to the deformation of organ and target contour and the calculation of dose distribution. In this paper, Symmetric is used in the study. Demons multi-resolution deformation registration method to extract the scope of CT images before and after radiotherapy. It is applied to PET images before radiotherapy and used to deform PET images before and after radiotherapy, so that the fusion of PET images before and after radiotherapy is more accurate, and compared with PET images after radiotherapy. By setting the threshold of SUV(standard uptake value in the image, the overlap rate between the outlines can be determined to obtain the high uptake area in the image. PET/CT data of 22 patients with non-small cell lung cancer (NSCLC) were analyzed and evaluated. Conclusion: based on the high overlap rate of high uptake region of Fludeoxyglucose fluorodeoxyglucose (Fludeoxyglucose) FDG. These results indicate that the location of the lesion is relatively stable before and after radiotherapy, and the residual tumors after X-ray irradiation (i.e. RTT) at the kilovolt or megovolt level are basically located in the target area before radiotherapy. The preliminary experimental results show that this study can be used to judge the response of target region to radiotherapy. The retrospective guidance radiotherapy plan can increase the radiation dose for the residual target area after radiotherapy, protect organs and tissues, and accurately radiotherapy.
【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】：R730.55;TP391.41

【参考文献】