节点文献

数字化女性盆腔放射治疗模型研究

Research on Digitized Radiotherapy Model of the Female Pelvic Cavity

【作者】 单锦露

【导师】 张绍祥;

【作者基本信息】 第三军医大学 , 人体解剖与组织胚胎学, 2008, 博士

【摘要】 肿瘤放射治疗是目前治疗恶性肿瘤的三大手段之一,其治疗方式是利用电离辐射对生物细胞的杀伤作用进行治疗。所以,作为治疗的各种放射线在杀灭肿瘤细胞的同时,又不可避免损伤部分正常的组织和器官。放射治疗的目标就是在尽可能提高肿瘤的受照射剂量的同时最大限度地保护肿瘤周围的正常组织,从而提高肿瘤的局部控制率,降低正常组织的放疗并发症。人体内器官的空间位置及辐射剂量的分布是无法通过仪表直接测量的,在对医疗仪器科学性、安全性、可操作性的判断过程中,不允许使用活体真人进行试验,因此必须为人类找到一个替身来研究器官在人体中的空间位置以及在医疗照射中器官剂量的分布来避免上述危害。中国数字化人体数据集的建立为放射治疗的剂量学计算提供了中国人人体模型基础,利用中国数字化人体数据集可以建立完整的人体数字化模型库,能更精确地对人体内各个器官的空间位置进行观测和对组织剂量分布进行数字模拟仿真计算,以便获得更合理的放疗方案和更精确的剂量分布,提高肿瘤的疗效,减少放疗并发症,并为制订符合中国人的放射物理模型提供了理论依据和技术平台。本实验以标本号为CVH-2的可视化女性人体数据集的原位女性盆腔连续横断层图像(髂嵴至坐骨结节下缘平面)为研究对象,利用Amira 4.1软件进行计算机图像重采样处理,生成多方位的连续薄层断面图像数据集,实现同一例标本的多方位断面图像的观测,在此基础上对常用照射野内的组织器官进行断层解剖学观测。在计算机上利用可视化软件,采用交互式手动分割和自动阈值分割相结合的数据提取方法,进行盆腔内放射治疗敏感器官的图像分割和三维重建,在盆腔器官三维重建的基础上,模拟盆腔常用的外照射射野,使得照射野所包含的组织在三维空间上可以直观地显示出来,并可在各个方位、角度上进行动态观察,讨论各种照射野所包含的解剖结构的空间位置关系以及常规照射野的一些挡铅方式存在的问题。在构建的女性盆腔内部器官结构三维解剖模型基础上,充分利用数字化人体数据的高清晰度、高分辨率、高识别率的优势,结合同一例数据标本的CT数据,探讨不同数据源与标准DICOM数据转换的方法,并将两种类型的数据进行了图像配准和融合,使低识别率的CT数据具有了高识别率的特性,构建临床上通用的女性盆腔数字化放射治疗模拟计算模型,以期实现在数字模型上对妇科肿瘤盆腔放射治疗中肿瘤靶体积和重要器官剂量的评估和放射治疗计划的优化设计,为建立完整的人体数字化辐射模型库方法进行初步探索。其研究成果对于盆腔恶性肿瘤放射治疗剂量的解析计算、临床治疗方案的选择与优化以及放射治疗计划质量保证(quality assurance,QA)具有重要的理论和临床意义。本研究的主要结果和结论如下:1.在女性数字化可视人体数据集盆腔薄层断面图像数据的基础上,利用Amira 4.1软件的重采样功能生成多方位的连续断面图像数据集,实现同一例标本的多方位断面图像观测,为开展数字解剖学教学与临床影像诊断等提供了全方位的断层解剖学资料。2.获得了女性盆腔放射治疗中多种照射野方位的连续断层图像,结合临床照射野的特点,探讨了相关照射野视窗范围内重要器官的断层解剖学结构特点与走行分布规律,为肿瘤影像诊断的准确性和放射治疗方式的正确选择、靶区的正确勾画提供断层解剖学依据。3.运用计算机三维重建可视化技术,实现了盆腔照射野的三维可视化,精确模拟了常规照射野的观察视窗,从多方位、多角度立体显示照射野范围内各器官之间的空间结构和毗邻关系,特别是靶器官和危险器官的毗邻关系,为女性盆腔外照射射野方式的选择和降低肿瘤放射治疗中放疗并发症发生的深入研究提供了立体形态学依据。4.利用虚拟断层技术和三维照射野视窗的虚拟现实仿真技术,为女性盆腔内器官结构的空间位置理解提供了更直观的显示方式,为照射野解剖学的深入研究提供了相应的技术平台。5.首次实现了数字化可视人体分割数据与标准DICOM3.0数据格式的转换,建立了一套符合临床放射治疗系统数据要求的标准DICOM3.0中国可视化人体数据源女性盆腔数据,为数字化可视人体应用于临床影像设备提供完整的技术方案和程序代码,有助于数字化可视人体数据的临床推广应用,为临床影像数据提供了精确而详细的辅助资料。6.对不同数据源的医学图像融合方法和融合效果评估进行了探讨,使同一例标本的CT数据源和数字化可视人体数据集的匹配度达到95%以上,并成功实现了同一例标本不同数据源的图像融合,使低分辨率的临床CT具有分割标识,为妇科肿瘤放射治疗过程中靶区勾画、治疗方案优化和剂量评估提供了具有重要器官分割特性的数字模拟计算模型。7.成功建立了平均CT值和具有分割特性的CT数据两种女性盆腔放疗模拟计算模型,所建模型符合DICOM3.0数据格式,可作为理想的中国人数字化放疗模拟计算模型,为研究更合理的放疗方案和更精确的剂量分布、寻求更加精确的剂量计算方法提供了理想的活体真人替代品,为建立完整的国人辐射人体数字化模拟计算模型库的方法进行初步探索。8.在临床放射治疗计划系统(TPS)上,利用建立的女性盆腔数字化放射治疗模拟计算模型实现了宫颈癌2D-RT、3D-CRT和IMRT剂量分布的模拟计算。在相应的剂量分布断面及三维重建图像上,对2D-RT、3D-CRT和IMRT在宫颈癌的剂量分布进行比较研究,探讨各种照射方式的优势和价值,为宫颈癌的临床放疗计划的设计提供了剂量学的参考依据。

【Abstract】 At present, radiotherapy is one of the three major techniques for treating cancer patients ,which causes cell injury when cancer patients were being exposed to the ionizing radiation. The radiation damage to normal tissues and organs of patients with cancer during the radiotherapy are unavoidable when radiation kills tumor cells. The desired therapeutic outcomes of radiotherapy are to maximally improve the tumor radiation dose and to maximally decrease the damage to ambient normal tissues and organs. The final goal of radiotherapy is to improve local control rate and to decrease post radiation complications of the cancer patients. The spatial position relation and the radiation doses distribution of human organ and tissues can not be directly measured and displayed by any apparatus. It is forbidden to use live human in the experiment of evaluating safety and maneuverability of medicinal apparatus. So a substitute for live human must be invented to study human’s organic architecture and radiation doses distribution in radiation instead of. The Chinese Visible Human (CVH) datasets provide the possibility of constructing digital human model and numerical data-base of the Chinese about radiotherapy dose calculation. We can accurately observe human’s organic architecture and calculate radiation doses simulation through digital human model, which would be the standard of the physical radiotherapy model from the Chinese.In this study, 386 continuous cross-sections of female pelvic cavity (between slices 1715 and 2100) were chosen from the second datasets of CVH. After being registered accurately, the cross-sections of female pelvic cavity data set were resampled to create the traverse, coronal and sagittal planes images data set with Amira4.1 software on PC. By using the visualizational software, the structures of radiosensitive organs were segmented by interactive and thresholds complex segmentation approach and were reconstructed to get female pelvic 3D model. Using the methods described above, we have reconstructed a numerical model of the inner organs of the female Visible Human pelvis. The important structures in the radiation field for female pelvic cavity including the pelvis, the uterus, the ovary, the ureter, the vesica, the internal iliac vein and artery, external iliac vein and artery were displayed clearly in false color by surface rendering reconstruction. In 3D view, the graphic modeling provides an overview of the pelvic structures and their relationships to each other and other anatomical regions. We can prune, enlarge, and roam the inner organs data in any orientation so that the inner structure of the radiation field for female pelvic cavity can be observed clearly. Then, the color images and the CT DICOM images of female pelvic cavity from CVH data set were registered and fused to construct a physical radiotherapy model of female pelvic cavity. Good image quality can be achieved by processing images of portal field and it is helpful to achieve high anatomy registration accuracy of the color images and CT images. Display of the fused images helps anatomy structure registration. Our rerults show that fusing images with anatomic images can help to obtain a full understanding of the portal field and peripheral structures and can be used as a routine clinical practice. This study have shown that spatial relationships between the anatomy structures are easier to understand when being visualized stereoscopically in 3D environments, and the physical radiotherapy model have therefore the potential to extend the radiotherapy planning process, giving an improved understanding of how the dose is distributed throughout the patient’s body. Also, this report presents a physical radiotherapy model for evaluating radiotherapy treatment plans, developed to improve the understanding of the spatial relationships between the patient anatomy and the calculated dose distribution. A physical radiotherapy model offers visualization through interactive volume rendering of radiotherapy dose distribution and computed tomography (CT) and surface and line rendering of radiotherapy structures such as target volumes and organs at risk. A physical radiotherapy model has been installed and networked in a hospital room used for the daily radiotherapy conferences, making stereoscopic viewing of treatment planning data for clinical cases possible.The main results and conclusions were as follows:1. The continuous cross-sections of female pelvic cavity chosen from the second datasets of CVH were resampled afterbeing registered accurately. The thin sectional anatomy data set from multidirectional planes were obtained on the same individual. It can help to promote image diagnosis and anatomical studies of digital visible human. 2. The cross-sectional anatomic characteristics and distribution of the internal structure were observed in the radiation field in female pelvic cavity. The cross-sectional anatomic characteristics of the radiation field in female pelvic cavity can be effectively used to recognize medical automatic images and get target therapy areas’image drawings in radiotherapy. It can help to choose optimal radiotherapy methods and to provide anatomical basis for the radiotherapy.3. By using visualizational software, the structures of radiosensitive organs were segmented by interactive and thresholds complex segmentation approach and were reconstructed to get female pelvic 3D model. We can prune, enlarge, and roam the inner organs data in any orientation so that the inner structure of the radiation field for female pelvic cavity can be observed clearly.4. Based on the digital medical image resampling techniques and virtual-reality technology, an intuitively understandable manner was provided to study the inner anatomic structure of the radiation field in female pelvic cavity. It provided a technology platform for research on anatomy of radiation field.5. This trial transformed the CVH color images to the DICOM files for the first time. The CVH color cross-sectional images of female pelvic cavity were transformed to DICOM3.0 images with Amira4.1 and Matlab software. A dataset of female pelvic DICOM files from CVH was set up to apply in radiation therapy system.6. This trial discussed the fusion method and efficiency evaluation for the multimodality medical image information. The pelvic cavity structures of Chinese Virtual Human dataset were segmented using image transparency method at first and each image was smoothed by eroding and dilating, and then fused successfully with CT of the same sample. The registration accuracy of CVH/CT was more than 95%. The digital radiotherapy model of female pelvic cavity with important organs segmented can well reflect the exposure doses of patients, so it is of important value to get target therapy areas’image drawings, forecast radiotherapy side effects, optimize radiotherapy plan and improve curative effect.7. This trial presented a digital physical radiotherapy model with the mean CT values and a model with the segmented CT data of female pelvic cavity for the first time. The color images and the CT DICOM images of female pelvic cavity from CVH were registered and fused to construct a physical radiotherapy model of female pelvic cavity. We can accurately observe human’s organic architecture and calculate radiation doses simulation in radiation therapy for patients with gynecological malignancies through the digital model, which would be the standard of the physical radiotherapy model from the Chinese. The methods of founding integrated numeric radiotherapy model were initially studied.8. The physical radiotherapy model of female pelvic cavity was imported in the radiotherapy planning system for the first time. The dose distribution of 2D-RT、3D-CRT and IMRT on the physical radiotherapy pelvic model from CVH2 was transferred and superimposed onto the sectional anatomy images and anthropomorphic pelvic phantom. The differences with target dose and normal tissue absorb dose by the DVHs were compared. Much more information about clinical dose distribution was obtained on anatomy images than that on CT images. The physical radiotherapy model on the TPS system has great value in the study of fine anatomy and clinical images. It provides technology platform for research and radiotherapy training.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络