改进的三维可视化用光线投射算法

把图像处理、光线投射与包围体技术有机结合，提出了一种提高成像质量和速度的三维可视化新方法。该方法利用物体空间的包围体算法来减少追踪光线的数量，加快了绘制速度。通过实际的医学胸部CT图像的三维重建实验，取得了较好的三维显示效果和速度，验证了改进的光线投射算法对胸部CT图像的快速三维可视化问题的有效性。     

一种改进的光线投射方法

光线投射算法是体绘制中的经典算法,但其绘制速度较慢。本文对传统的光线投射算法中等间距重采样进行了改进,引入包围盒方法,采用变步长的采样方法,减少冗余数据量和投射光线数量,优化重采样过程,提高采样效率。医学图像可视化实验表明,改进方法能够在保证图像质量的同时,提高绘制速度。     

工业过程断层图像的三维动态可视化

提出了一种针对工业过程断层图像的三维动态可视化方法,可用于对工业过程的辅助监控。该方法采用基于光线投射的体可视化技术,已在MITK(Medical Imaging ToolKit,一个用于医学影像处理与分析的C++类库)中实现。该方法使用不同的颜色和阻光度系数来区分反应容器或管道中的不同物质,从而为容器或管道中不同物质的混合反应过程提供一个动态的显示。实验结果证明该方法是可行的,并且其性能也是可接受的,若再辅之以体绘制算法的硬件加速,该方法可用于实时的工业断层成像系统中。     

基于光线投射的锥束CT正投影算法研究

计算机断层成像（computed tomography,CT）作为一种现代成像技术已经被广泛应用于医学诊断、工业无损检测等领域。在研究CT重建算法的过程中,正投影模拟数据是必不可少的一部分。本文提出一种基于光线投射算法的圆轨迹锥束扫描模式下的正投影算法,并结合CUDA（compute unified device architecture,统一计算设备架构）技术,实现了GPU（graphics processing unit,图形处理器）加速正投影计算。通过与传统的正投影算法相比,本文算法在投影图像质量上有一定的提高,并具有更高的计算效率。     

基于多等值面的工业CT三维显示与测量

结合大型工业CT中对缺陷检测、参数测量、空间密度分布检测等工程应用要求,研究了应用于工业CT缺陷检测的表面绘制可视化检查技术。基于Marching Cube算法构造等值面的原理,分析了分析工业CT检测对象特点和先验知识,设计了工业CT工件模型化区域多等值面可视化检查的方法,在OpenGL面绘制中实现了基于深度缓存的交互式标定测量的原理和算法,为工业CT体数据的三维可视化检查提供有力的手段。     

显微三维分析软件设计与开发

论述了显微三维分析软件系统的组织结构和程序流程; 针对因离焦光线的干扰而产生的模糊光学断层图像序列, 基于断层内相邻像素和断层间相邻像素的相关信息, 采用改进的最大期望算法对光学断层图像序列进行去模糊处理; 对去模糊处理后的三维光学断层图像序列, 介绍了一种基于表面点绘制的三维数据场表面重建反走样方法, 加快了重建速度, 同时具有较好的显示效果; 三维体绘制时, 提出了多维半自动阻光度转换函数解决方案, 清晰地显示出了物体内部的细节变化; 采用基于体素的方法计算重建后的物体的表面积和体积; 通过三维分割和标记算法, 实现了在三维空间内选择感兴趣目标的操作。     

三维锥形束CT解析重建算法发展综述

同二维的扇束、平行束相比,三维锥束CT需要的扫描时间更短,可以获得更好的空间分辨率和更高的射线利用效率。虽然锥束CT的解析重建算法在数学计算上比较复杂,由于其运算量较大,工程实现起来也有一定的困难,但是随着近几年硬件和算法的快速发展,三维锥束CT变得越来越有希望,医用及工业CT正向着中等甚至大锥角三维锥束CT过渡。鉴于其巨大的实用背景,本文对近些年三维锥束解析重建算法的发展做了一个回顾,尤其是针对长物体问题的算法及短物体问题的算法进行的研究,并对各类算法作了比较和讨论,最后对三维锥形束CT解析重建算法理论的发展进行了展望。     

体层次迭代发射算法

本文介绍体层次迭代发射算法 ,它用于虚拟现实图形演示的体放射法。基于体单元对光线的发射、吸收、漫射、反射和散射基础上的体放射法 ,适用于复杂景物的生成和演示 ,在真实感图形演示中具有重要地位。本文首先介绍体放射法的主要技术细节 ,如光线计算 ,体放射方程 ,光线生成参数 ,接着说明体层次发射算法的机理 ,最后给出了本算法生成的实例。同时对算法还做了评估。     

新的超分辨CT扫描模式和图像重建方法

针对通常工业CT转台特点,提出了一种新的基于扇束等距采样的超分辨CT扫描模式和相应的重建算法。分别向左和右偏移CT旋转中心1/4探测器宽度,得到的两组相叠采样的投影数据。通过求解一组具有双对角矩阵的线性方程组,得到高分辨的CT虚拟投影数据,进而由扇束滤波反投影算法重建出高分辨的CT图像。数值模拟结果验证了本文方法的有效性。     

双能DR物质识别算法在CT成像系统中的应用

本文介绍一种适用于计算机断层成像（CT）的双能透视（DR）物质识别算法，此算法基于单能CT重建图像，通过对CT图像的分割及对断层几何信息的提取，分块重建材料有效原子序数及电子密度的分布。结合各种扫描轨迹的CT成像系统，可以实现有效的物质材料识别。同时，对比于传统的双能CT方法，本方法结合单能CT重建图像，改善了双能DR识别效果，能够实现较为精确的物质识别，在安全检查等应用领域有着现实意义。     

Volumetric visualization of head and neck CT data for treatment planning.

PURPOSE: To demonstrate the utility of volume rendering, an alternative visualization technique to surface rendering, in the practice of CT based radiotherapy planning for the head and neck. METHODS AND MATERIALS: Rendo-avs, a volume visualization tool developed at the University of Chicago, was used to volume render head and neck CT scans from two cases. Rendo-avs is a volume rendering tool operating within the graphical user interface environment of AVS (Application Visualization System). Users adjust the opacity of various tissues by defining the opacity transfer function (OTF), a function which preclassifies voxels by opacity prior to rendering. By defining the opacity map (OTF), the user selectively enhances and suppresses structures of various intensity. Additional graphics tools are available within the AVS network, allowing for the manipulation of perspective, field of view, data orientation. Users may draw directly on volume rendered images, create a partial surface, and thereby correlate objects in the 3D scene to points on original axial slices. Information in volume rendered images is mapped into the original CT slices via a Z buffer, which contains the depth information (Z coordinate) for each pixel in the rendered view. Locally developed software was used to project conventionally designed GTV contours onto volume rendered images. RESULTS: The lymph nodes, salivary glands, vessels, and airway are visualized in detail without prior manual segmentation. Volume rendering can be used to explore the finer anatomic structures that appear on consecutive axial slices as "points." Rendo-avs allowed for acceptable interactivity, with a processing time of approximately 5 seconds per 256 x 256 pixel output image. CONCLUSIONS: Volume rendering is a useful alternative to surface rendering, offering high-quality visualization, 3D anatomic delineation, and time savings to the user, due to the elimination of manual segmentation as a preprocessing step. Volume rendered images can be merged with conventional treatment planning images to add anatomic information to the treatment planning process.     

基于PC机的大规模体数据场实时可视化

为了能够快速可视化大规模体数据场,以往的研究主要集中在基于高端图形工作站和计算机集群上的并行算法。近年来个人计算机的性能大幅提高,特别是可编程图形硬件飞速发展,利用PC机可视化大规模体数据场成为可能。本文探讨了基于PC机可视化大规模体数据场的可行性,总结了体绘制软件和硬件加速算法,特别是基于可编程图形硬件的体绘制加速算法,并指出了进一步的研究方向,为十亿字节以上的大规模体数据场的实时高质量可视化提供了一个可借鉴的解决方案。     

首都师范大学检测成像实验室CT研究进展

CT检测成像技术包括数据采集、图像重建、图像应用三个层面。本文简要介绍了近年来首都师范大学检测成像实验室在这三个层面技术研究方面所取得的进展。具体包括多种新的扫描模式和相应重建算法、X射线能量谱的研究、基于GPU的多种解析重建算法和迭代重建算法的加速实现方法、多种伪影形成机理的研究及有效的伪影校正方法、基于GPU实现的体绘制算法、基于这些研究成果的算法平台和软件研发以及成果推广应用情况等。本文目的在于与本领域专家交流相关问题的国内外研究状况及本实验室的研究进展,促进研究的进一步深入。     

Online volume rendering of incrementally accumulated LSCEM images for superficial oral cancer detection

Laser scanning confocal endomicroscope (LSCEM) has emerged as an imaging modality which provides non-invasive, in vivo imaging of biological tissue on a microscopic scale. Scientific visualizations for LSCEM datasets captured by current imaging systems require these datasets to be fully acquired and brought to a separate rendering machine. To extend the features and capabilities of this modality, we propose a system which is capable of performing realtime visualization of LSCEM datasets. Using field-programmable gate arrays, our system performs three tasks in parallel: (1) automated control of dataset acquisition; (2) imaging-rendering system synchronization; and (3) realtime volume rendering of dynamic datasets. Through fusion of LSCEM imaging and volume rendering processes, acquired datasets can be visualized in realtime to provide an immediate perception of the image quality and biological conditions of the subject, further assisting in realtime cancer diagnosis. Subsequently, the imaging procedure can be improved for more accurate diagnosis and reduce the need for repeating the process due to unsatisfactory datasets.     

Comparison of volume-rendered and surface-rendered MR colonography

In the United States and Europe, colorectal cancer is the second leading cause of cancer-related deaths. It is well known that colorectal carcinomas may originate from preexisting adenomas. For the visualization of colorectal cancer and other pathologic changes such as polyps, two 3D methods (volume-rendering (VR) and surface-rendering (SR)) in MR colonography were compared in our study. MR colonography was carried out in 17 patients on a 1.5 T MR scanner using a 10 mmolar gadolinium water solution enema. Coronal as well as rotated VR and SR views were compared in order to examine the technical quality (TQ) of the visualization model and grade of confidence (GC) in the pathological findings. Colonoscopic findings revealed 8 colorectal carcinoma, 10 patients with polyps, 4 diverticular disease, and 2 with redundant bowel loops. Based on a total of 248 colonic segments for both visualization methods, volume rendering were significantly superior to surface rendering for both, TQ (p<0.0001) and GC (p<0.0001). Volume rendering and surface rendering were not dependent on individual colon segments (p=0.13 for TQ and p=0.18 for GC) or on image rotation (p=0.06 for TQ and p=0.062 for GC). It is also independent of the type of pathology (p=0.31 for TQ and p=0.42 for GC) and the reviewers (p=0.62 for TQ and p=0.88 for GC). This indicates, that for the purpose of interpreting the technical quality and pathological findings, volume rendering is superior to surface rendering in MR colonography. Volume rendering could be used as an 3D visualization tool, enabling MR colonography examinations to be completed sooner in cases where colon distension is sufficient, and it would also provide an overview of potential mass lesions.     

Prospects for proton therapy in carcinoma of the cervix.

Carcinoma of the cervix remains a serious problem worldwide. The results of radiation therapy varies from institution to institution, with actuarial 5-year survival rates for Stage IIIB disease of between 28 and 60%. The inferior survival figures can be ascribed to poor local control due to inadequate dosage, partly due to the non-use of intracavitary therapy. Results from the literature were reviewed to assess the optimal dose required to sterilize the primary, para-metrial and nodal disease within the pelvis. The aim of this study is to assess whether the target volume could be reduced to allow an increased dose to be delivered. Protons, in view of their physical characteristics, would aid this objective. From CT scans the minimal target volume was determined, and treatment plans for a 200 MeV proton beam were evaluated. It appears that a volume reduction of up to 60% in the target volume may be achieved with proton therapy when compared to the usual target volume achievable with photon therapy, and thus a 20% dose increment may be reasonable. This increased dose may translate to a 40% improvement in the local control rate without an expected increase in complications. An improved local control rate may lead to improved survival. A compatible technique of well-fractionated high dose rate intracavitary therapy (ICT) based on an indwelling intra-uterine tube is offered that will allow up to 10 fractions of ICT with only a single anaesthetic, to be integrated with a similar number of fractions of proton therapy.