基于工业CT断层图像的三维可视化

工业CT图像三维可视化能够对工业构件提供真实、直观的反映。体绘制技术可以显示工业CT三维数据的整体特征和内部细节信息。根据光线投射算法的特点，采用对原始数据场进行最大熵原则的预处理的方法，加快了绘制速度，在一定程度上改进了光线投射算法，取得了较好的显示效果。     

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。     

锥束/多层螺旋CT重建算法在工业CT中的应用及对运算速度的研究

目前临床应用的锥束/多层螺旋CT主要采用螺旋Feldkamp算法进行图像重建,体数据的重建需要消耗大量的时间,锥束/多层螺旋CT的图像重建一直是图像重建中的关键问题和衡量CT系统的重要指标之一。本文分析了几种快速重建算法的特点,并从算法结构、实现技巧及代码优化等方面论述了图像重建的方法,计算模拟和实验结果显示了这些方法可有效提高图像重建的速度。     

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。     

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。     

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。     

组织切片图像的可视化技术及应用

目的:研究和开发序列组织切片图像的三维显示技术。方法:首先利用序列切片中上下层切片相关点的对应关系,用最小二乘方法实现上下层切片图像的几何变换配准,进而用序列图像插值算法完成组织切片图像的体数据重建。结果:在对重点显示目标分类的基础上,用基于物空间SHEAR-WARP快速直接体视方法对体数据进行三维显示,组成一套医学图像可视化系统。结论:该系统能对胆管癌等序列切片图像进行三维重建显示。     

放疗中基于双任务循环网络的单投影重建胸部CT方法研究北大核心CSCD

目的基于cycleGAN建立双任务循环网络,实现基于单角度投影的三维CT合成模型用于胸部肿瘤的自适应放疗,并评估图像生成质量和剂量精度。方法收集南京医科大学附属常州第二人民医院收治的胸部肿瘤容积弧形调强放射治疗(VMAT)患者45例,并筛选公共数据集图像991例作为预训练数据集,使用ASTRA算法获取多角度投影图像。公共数据集划分为训练集800例、验证集160例和测试集31例,医院患者数据划分为训练集40例、测试集5例。分别输入合成CT模型和多角度投影预测模型,实现双任务的训练。最终测试仅使用合成CT模型获取预测CT图像进行图像生成质量[包括平均绝对误差(MAE)、峰值信噪比(PSNR)和结构相似性指数(SSIM)]以及剂量学评估。结果合成CT的图像质量评估指标显示,图像合成精度较高,MAE为0.05±0.01,PSNR为19.08±1.69,SSIM为0.75±0.04。基于合成CT计算的放疗剂量分布也与真实分布接近,3%/3 mm标准下γ通过率平均为93.1%。结论由cycleGAN修改的双任务循环网络,可以快速准确地从单角度投影预测三维CT,可以应用于胸部肿瘤自适应放疗的工作流程。图像生成质量和剂量学的评估均显示合成CT质量可以满足放疗的临床要求。     

ASIR-V重建算法在胸部低剂量CT诊断肺结节中的临床价值

目的:探讨胸部低剂量CT结合基于多模型的迭代重建算法（ASIR-V）在肺结节诊断中的临床价值。方法:收集行Revolution CT胸部平扫的肺结节患者40例。患者首次检查时先采用常规剂量胸部CT扫描,预设噪声指数为14 HU,图像重建采用滤波反投影法（FBP）;若发现肺结节者,于肺结节局部层面进行低剂量靶扫描,预设噪声指数为24 HU,图像采用60%ASIR-V算法进行重建。记录两种扫描模式下肺结节的检出数,测量肺结节的大小、CT值、噪声值以及肺组织CT值和噪声值,并计算肺结节的信噪比（SNR）和对比噪声比（CNR）。两名有经验的放射科医师采用5分法对两种扫描模式下整体图像质量以及部分肺结节的特殊形态学征象进行主观评分。记录两种扫描模式中患者的CT容积剂量指数(CTDIvol)、剂量长度乘积(DLP),并计算有效剂量(ED)。结果:常规剂量扫描ED为(3.20±1.14) mSv,低剂量扫描ED为(1.64±0.29) mSv,辐射剂量减低约48.8%。常规剂量扫描共检出108个肺结节,低剂量扫描共检出107个肺结节,低剂量CT扫描肺结节检出率为99.07%。低剂量60%ASIR-V图像和常规剂量FBP图像在肺结节的大小、CT值、SD值、SNR及CNR方面差异均无统计学意义（P＞0.05）。两名医师对两组图像的主观评分差异亦无统计学意义（P＞0.05）。结论:在低剂量CT扫描中,运用60% ASIR-V算法对于肺结节的检出、肺结节形态学特征的显示与常规剂量FBP图像质量差异不大,但其辐射剂量明显降低。     

用于颅颌面种植外科的三维立体可视化系统

目的论文对在个人PC机上对颅颌面医学图像的可视化进行研究,开发出用于颅颌面种植外科的CT断层图像三维可视化系统。方法应用3D纹理映射的方法对颅颌面CT断层图像进行三维重建,并得到其XYZ方向上的切面图像。然后利用边缘检测的方法得到眼眶的位置,并在眼眶位置进行扇形切割,得到18幅扇形切割图像。结果该系统可以帮助医生进行手术前的骨质、骨量分析,设计种植手术的过程及模拟、预测手术的种植效果。结论建立的颅颌面种植外科三维可视化系统以其硬件配置合理、软件设计新颖、多维视角、便捷快速精确等为特色。具有较强的应用价值。     

改进互信息与图像金字塔相结合的二维与三维图像配准方法研究

目的 探讨改进互信息与图像金字塔相结合的二维-三维配准方法的价值。方法 将3次B样条曲线的连续图像表示与Parzen直方图估计融合到该算法中,以胸腔作为研究对象,通过数字重建放射影像生成的正交位的模拟X线图和与其本身进行一定变换后的图像进行配准实验,观察配准精度和时间。结果 经过50组对照试验得出本方法相较于传统配准方法在x、y轴向上的位移精度分别提高了53.39%、21.33%,配准时间缩短了91.93%。相较于近几年的改进算法在x、y轴向上的位移精度分别提高了17.65%、13.79%,并将配准时间进一步提高了19.64%。结论 该方法可以有效提高二维-三维图像的配准精度和效率,且均符合手术过程中图像配准2mm以内的要求。该方法的高效、准确为临床诊断和放疗自动化提供了有利的信息,同时也为肿瘤位置误差校正和医用机械臂自动摆位奠定了基础。     

A novel 3D volumetric voxel registration technique for volume-view-guided image registration of multiple imaging modalities

PURPOSE: To provide more clinically useful image registration with improved accuracy and reduced time, a novel technique of three-dimensional (3D) volumetric voxel registration of multimodality images is developed. METHODS AND MATERIALS: This technique can register up to four concurrent images from multi-modalities with volume view guidance. Various visualization effects can be applied, facilitating global and internal voxel registration. Fourteen computed tomography/magnetic resonance (CT/MR) image sets and two computed tomography/positron emission tomography (CT/PET) image sets are used. For comparison, an automatic registration technique using maximization of mutual information (MMI) and a three-orthogonal-planar (3P) registration technique are used. RESULTS: Visually sensitive registration criteria for CT/MR and CT/PET have been established, including the homogeneity of color distribution. Based on the registration results of 14 CT/MR images, the 3D voxel technique is in excellent agreement with the automatic MMI technique and is indicatory of a global positioning error (defined as the means and standard deviations of the error distribution) using the 3P pixel technique: 1.8 degrees +/- 1.2 degrees in rotation and 2.0 +/- 1.3 (voxel unit) in translation. To the best of our knowledge, this is the first time that such positioning error has been addressed. CONCLUSION: This novel 3D voxel technique establishes volume-view-guided image registration of up to four modalities. It improves registration accuracy with reduced time, compared with the 3P pixel technique. This article suggests that any interactive and automatic registration should be safe-guarded using the 3D voxel technique.     

Computation of digitally reconstructed radiographs for use in radiotherapy treatment design

The increasing use of 3-dimensional radiotherapy treatment design has created greater reliance on methods for computing images from CT data which correspond to the conventional simulation film. These images, known as computed or digitally reconstructed radiographs, serve as reference images for verification of computer-designed treatments. Used with software that registers graphic overlays of target and anatomic structures, digitally reconstructed radiographs are also valuable tools for designing portal shape. We have developed radiograph reconstruction software that takes full advantage of the contrast and spatial detail inherent in the original CT data. This goal has been achieved by using a ray casting algorithm which explicitly takes into account every intersected voxel, and a heuristic approach for approximating the images that would result from purely photoelectric or Compton interactions. The software also offers utilities to superimpose outlines of anatomic structures, field edges, beam crosshairs, and linear scales on digitally reconstructed radiographs. The pixel size of the computed image can be controlled, and several methods of interslice interpolation are offered. The software is written in modular format in the C language, and can stand alone or interface with other treatment planning software.     

CT Colonography - Towards Applications for Colorectal Cancer Screening

Three-dimensional (3D) imaging of the large intestine is globally called computed tomography colonography(CTC). CTC has been intensively investigated for application in colorectal cancer screening in Western countriesand with the advent of multi-slice CT (MSCT), which provides effective high resolution in 3D CT images, thediagnostic use of CT for colorectal lesions has become a concept widely accepted throughout the world. Computeraideddetection (CAD) for colorectal polyps using digital CT image data and digital pre-processing are alsoadvancing in the West. Compared with colonoscopy, which depends largely on the skill of the performer, CTCproduces objective and reproducible diagnostic images and presents a high probability of standardizingexamination protocols. Development of effective systems for screening colorectal lesions is expected, leveragingthe excellent processing capability of MSCT to enhance 3D visualization and allow efficient detection.     

A technique for optimization of digitally reconstructed radiographs of the chest in virtual simulation

INTRODUCTION: Virtual simulation (VS) of radiotherapy uses CT data. Digitally reconstructed radiographs (DRRs) are a critical element of this process, and the quality of these images is frequently suboptimal. We present techniques to improve DRR quality for clinical purposes. The results of two approaches to DRR optimization are presented. METHODS AND MATERIALS: One approach to DRR optimization is to use traditional radiographs as a guide and to adjust the algorithm parameters based on image and objective contrast to produce images that more closely resemble traditional radiographs (Method 1). Another approach is to focus on the visibility of specific anatomic structures. Using this method, two DRR images are optimized manually by interactively adjusting reconstruction parameters, then they are combined into a single composite image (Method 2). DRRs for the chest region, generated using both methods, were evaluated by clinical staff based on usability for treatment verification and field definition. RESULTS: Using Method 1, the resulting DRRs more closely resembled traditional radiographs. This technique allows DRR quality to be improved with little user interaction. These DRRs are generally adequate for clinical use, but not optimal for sites such as the chest. Images generated using Method 2 were considered clinically superior in terms of visibility of specific anatomic structures. These images also compare well with traditional radiographs, although they show an increased contrast level between bone and lower density structures. CONCLUSION: Both Methods 1 and 2 can be used to improve DRR quality for clinical purposes. For the chest region, the additional effort required by Method 2 to achieve a more detailed image appears justified.     

A feasibility study for image guided radiotherapy using low dose, high speed, cone beam X-ray volumetric imaging.

BACKGROUND AND PURPOSE: Image Guidance of patient set-up for radiotherapy can be achieved by acquiring X-ray volumetric images (XVI) with Elekta Synergy and registering these to the planning CT scan. This enables full 3D registration of structures from similar 3D imaging modalities and offers superior image quality, rotational set-up information and a large field of view. This study uses the head section of the Rando phantom to demonstrate a new paradigm of faster, lower dose XVI that still allows registration to high precision. MATERIALS AND METHODS: One high exposure XVI scan and one low exposure XVI scan were performed with a Rando Head Phantom. The second scan was used to simulate ultra low dose, fast acquisition, full and half scans by discarding a large number of projections before reconstruction. Dose measurements were performed using Thermo Luminescent Dosimeters (TLD) and an ion chamber. The reconstructed XVI scans were automatically registered with a helical CT scan of the Rando Head using the volumetric, grey-level, cross-correlation algorithm implemented in the Syntegra software package (Philips Medical Systems). Reproducibility of the registration process was investigated. RESULTS: In both XVI scans the body surface, bone-tissue and tissue air interfaces were clearly visible. Although the subjective image quality of the low dose cone beam scan was reduced, registration of both cone beam scans with the planning CT scan agreed within 0.1 mm and 0.1 degrees . Dose to the patient was reduced from 28mGy to less than 1mGy and the equivalent scan speed reduced to one minute or less. CONCLUSIONS: Automatic 3D registration of high speed, ultra low dose XVI scans with the planning CT scan can be used for precision 3D patient set-up verification/image guidance on a daily basis with out loss of accuracy when compared to higher dose XVI scans.