CT Dose Reduction by Fast Iterative Algorithms

通过快速迭代算法减少 CT 剂量

• 批准号: 7623952
• 负责人: Jeffrey Brokish
• 金额: $ 35.3万
• 依托单位: INSTARECON, INC.
• 依托单位国家: 美国
• 财政年份: 2005
• 资助国家: 美国
• 起止时间: 2005-08-01 至 2011-10-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/7623952
• 关键词: Acceleration; Address; Adoption; Adult; Algorithms; Architecture; Artificial cardiac pacemaker; Arts; Attention; Back; Benchmarking; Calculi; Cardiac; Cells; Characteristics; Chicago; Childhood; Clinical Data; Code; Collaborations; Complex; Computational Technique; Computer software; Coupling; Data; Development; Diagnostic; Diagnostic Imaging; Dose; Evaluation; Fluoroscopy; Foxes; Future; Goals; Grant; Healthcare; Heart Diseases; High Performance Computing; Illinois; Image; Implant; Legal patent; Letters; Literature; Manufacturer Name; Maps; Marketing; Measures; Metals; Methodology; Methods; Modeling; Morphologic artifacts; Noise; Non-linear Models; Operative Surgical Procedures; Output; Patients; Performance; Phase; Physics; Play; Plug-in; Process; Prosthesis; Public Health; Publishing; Radiation; Radiology Specialty; Relative (related person); Research; Research Project Grants; Resolution; Resources; Retinal Cone; Risk; Role; Running; Sales; Scanning; Scheme; Screening procedure; Side; Slice; Small Business Innovation Research Grant; Solutions; Source; Speed; Statistical Models; System; Systems Integration; Technology; Testing; Three-Dimensional Imaging; Time; Translating; United States National Institutes of Health; Universities; Upper arm; Variant; Work; X-Ray Computed Tomography; base; clinical application; clinical practice; commercialization; data acquisition; design; detector; image reconstruction; improved; in vivo; interest; lung cancer screening; meetings; member; new technology; next generation; novel; parallel computing; parallel processing; physical model; professor; prototype; radiologist; reconstruction; statistics; success; three dimensional structure; tomography

DESCRIPTION (provided by applicant): With the increased use of x-ray CT, the development of a market for CT screening exams, and the imaging of younger patients, there is a growing concern about the public health risk caused by the radiation dose delivered by x-ray CT. The reduction of this dose has therefore taken on increased importance. The objective of this work is to develop a novel computationally-based approach to the reduction of patient x-ray dose in diagnostic CT scanners. The approach will use iterative algorithms for the image formation, which can produce high-quality images from low-dose data by incorporating detailed models of the physics and statistics of the data acquisition process. To date, such iterative algorithms have been little used in practice owing to their high computational complexity. This problem will be solved by using revolutionary fast algorithms for the backprojection and reprojection steps in the iterative algorithm. These fast algorithms were developed and patented at the University of Illinois by members of the project research team and collaborators, and further developed at InstaRecon, Inc. Using this technology, speed-up factors of 10x - 50x have been achieved in software prototypes. Phase I developed fast statistical and physics-based iterative algorithms for reduced-dose and reduced artifact high-precision tomography for the 2D fan-beam imaging geometry. In Phase II, the methodology and algorithms will be extended to the dominant imaging geometries in modern multi-detector-row diagnostic scanners: helical multislice, conebeam with a circular source trajectory, and helical conebeam. The success of Phase I suggests an acceleration of the iterative algorithm by a factor of 10 or better compared to previous, conventional implementations. This large algorithmic acceleration will be further augmented by implementing the algorithms on a parallel computing platform. The resulting prototype reconstruction system will match the throughput of current CT scanners, while providing dose and artifact reduction. Significant attention will be devoted to thorough testing and quantitative characterization of the speed and image quality of the new dose reduction technology. Benefits of the new technology will include superior low-dose performance in dose-critical applications such as pediatric, screening for lung cancer or heart disease, and interventional imaging. Additionally, it will offer significant improvement in diagnostic quality of CT scans of large patients and of patients with prosthetic implants or cardiac pacemakers. The algorithmic speedup allows for the system to run on a modest hardware platform, making this technology attractive for adoption by scanner manufacturers. This project promises to revolutionize CT as we know it, by making iterative algorithm-based dose and artifact reduction feasible for the first time.

描述（由申请人提供）：随着X射线CT使用的增加、CT筛查检查市场的发展以及年轻患者的成像，人们越来越关注X射线CT辐射剂量引起的公共健康风险。因此，减少这一剂量变得越来越重要。这项工作的目的是开发一种新的基于计算的方法来减少诊断CT扫描仪中的患者X射线剂量。该方法将使用迭代算法进行成像，通过结合数据采集过程的物理和统计的详细模型，可以从低剂量数据中生成高质量的图像。迄今为止，这种迭代算法由于其高计算复杂度而很少在实际中使用。这个问题将通过使用革命性的快速算法的反投影和重投影步骤中的迭代算法来解决。这些快速算法由项目研究团队成员和合作者在伊利诺伊大学开发并获得专利，并在InstaRecon，Inc.进一步开发。使用这种技术，在软件原型中实现了10倍-50倍的加速系数。第一阶段开发了快速统计和基于物理的迭代算法，用于2D扇束成像几何结构的减少剂量和减少伪影的高精度断层扫描。在第二阶段，方法和算法将扩展到现代多探测器行诊断扫描仪中的主要成像几何形状：螺旋多层、具有圆形源轨迹的锥束和螺旋锥束。第一阶段的成功表明，与之前的传统实现相比，迭代算法的速度加快了10倍或更好。通过在并行计算平台上实现算法，将进一步增强这种大的算法加速。由此产生的原型重建系统将匹配当前CT扫描仪的吞吐量，同时提供剂量和伪影减少。将对新的剂量减少技术的速度和图像质量进行全面测试和定量表征。这项新技术的好处将包括在剂量关键型应用中的上级低剂量性能，如儿科、肺癌或心脏病筛查和介入成像。此外，它将显著提高大型患者和植入假体或心脏起搏器患者的CT扫描诊断质量。算法加速允许系统在适度的硬件平台上运行，使这项技术对扫描仪制造商的采用具有吸引力。该项目有望彻底改变CT，因为我们知道它，使迭代算法为基础的剂量和伪影减少可行的第一次。