Fourier-based Methods for Image Reconstruction in PET

基于傅立叶的 PET 图像重建方法

基本信息

批准号：
8235078
负责人：
SAMUEL MATEJ
金额：
$ 39.05万
依托单位：
UNIVERSITY OF PENNSYLVANIA
依托单位国家：
美国
项目类别：
财政年份：
2002
资助国家：
美国
起止时间：
2002-04-01 至 2015-03-31
项目状态：
已结题

项目摘要

DESCRIPTION (provided by applicant): The overall objective of this project is to improve the quality of images obtained by positron emission tomography (PET) for human studies in clinical nuclear medicine. This will be done by developing a fast and accurate computer method for generating images from basic photon-count data acquired by PET scanners having detectors with time-of-flight (TOF) capability. Data from TOF-PET systems contain additional information that permits better spatial localization of coincidence events, compared to conventional (non-TOF) scanners. TOF-PET scanners have been shown to yield significantly better images, especially for large patients; however, the full benefit of TOF has not yet been achieved in the clinical environment due to the relatively slow reconstruction techniques available for TOF-PET data. The proposed work involves the development and testing of an iterative computer method for statistical image reconstruction that makes specific use of the localized nature of TOF-PET data. The method is called DIRECT, short for Direct Image Reconstruction for TOF, and it involves grouping the TOF-PET data based on a novel combination of angular intervals in data space and voxel-like partitions in image space. The hypothesis is that this method will achieve high quantitative accuracy, combined with high computational efficiency, which is critical for obtaining these quantitative images in the clinical environment. High computational efficiency is needed for human studies, since a large amount of data is collected, the image space is sampled on a fine grid, and many iterations are required to accurately recover the activity levels at all locations in the body. Multi-frame studies (e.g., dual time point imaging, dynamic studies) involve multiple runs of the image reconstruction process for which a fast reconstruction technique is essential. In the DIRECT method, the novel grouping of TOF-PET data leads to high efficiency for many of the reconstruction operations; in particular, this grouping enables the operations of forward-projection and back-projection to be done using efficient Fourier-based methods. Achievement of high accuracy combined with high computational efficiency would be a significant step towards realizing the full potential of TOF-PET in the clinical environment, since in current practice performance is compromised for efficiency in order to complete routine whole-body studies in a practical time. Specific aim 1 is designed to formulate, implement, and investigate the DIRECT method for iterative image reconstruction in TOF-PET, focusing on the core components of the method. Specific aim 2 is designed to formulate, implement, and investigate those components of the DIRECT method that involve compensation for the non-ideal characteristics of measured data, including attenuation, scatter, randoms, and detector normalization. Specific aim 3 involves evaluation of the performance of DIRECT in comparison with other TOF-PET reconstruction techniques. PUBLIC HEALTH RELEVANCE: Positron emission tomography is now well established as a valuable imaging tool for the diagnosis of cancer and other diseases and for the planning and monitoring of treatment. The proposed work involves new methods for computer processing of data from a technically advanced generation of PET scanner; the new computer methods are designed to enable these scanners to reach their full potential, leading to improved accuracy of PET images. The proposed work is relevant to public health, since an improvement in the accuracy of PET images would lead to more accurate diagnosis of cancer and other diseases, more accurate planning of treatment, and more accurate monitoring of the response to therapy, leading in turn to better patient outcomes.

描述（由申请人提供）：该项目的总体目标是提高临床核医学人体研究中正电子发射断层扫描（PET）获得的图像质量。这将通过开发一种快速、准确的计算机方法来实现，该方法可根据具有飞行时间 (TOF) 功能的探测器的 PET 扫描仪采集的基本光子计数数据生成图像。与传统（非 TOF）扫描仪相比，TOF-PET 系统的数据包含额外的信息，可以更好地空间定位符合事件。 TOF-PET 扫描仪已被证明可以产生明显更好的图像，特别是对于体型较大的患者；然而，由于 TOF-PET 数据的重建技术相对较慢，TOF 的全部优势尚未在临床环境中实现。拟议的工作涉及开发和测试用于统计图像重建的迭代计算机方法，该方法专门利用 TOF-PET 数据的局部性质。该方法称为 DIRECT，是 TOF 直接图像重建的缩写，它涉及基于数据空间中的角间隔和图像空间中的体素类分区的新颖组合对 TOF-PET 数据进行分组。假设该方法将实现高定量精度和高计算效率，这对于在临床环境中获得这些定量图像至关重要。人体研究需要高计算效率，因为需要收集大量数据，在精细网格上对图像空间进行采样，并且需要多次迭代才能准确恢复身体所有位置的活动水平。多帧研究（例如，双时间点成像、动态研究）涉及图像重建过程的多次运行，为此快速重建技术至关重要。在 DIRECT 方法中，TOF-PET 数据的新颖分组可以提高许多重建操作的效率；特别是，这种分组使得可以使用高效的基于傅里叶的方法来完成前向投影和后向投影的操作。实现高精度与高计算效率相结合将是在临床环境中实现 TOF-PET 全部潜力的重要一步，因为在当前的实践中，为了在实际时间内完成常规全身研究，性能会因效率而受到影响。具体目标 1 旨在制定、实施和研究 TOF-PET 中迭代图像重建的 DIRECT 方法，重点关注该方法的核心组成部分。具体目标 2 旨在制定、实施和研究 DIRECT 方法的那些组件，这些组件涉及对测量数据的非理想特性的补偿，包括衰减、散射、随机性和探测器归一化。具体目标 3 涉及评估 DIRECT 与其他 TOF-PET 重建技术的性能比较。公共健康相关性：正电子发射断层扫描现已成为一种有价值的成像工具，可用于诊断癌症和其他疾病以及用于规划和监测治疗。拟议的工作涉及对来自技术先进的一代 PET 扫描仪的数据进行计算机处理的新方法；新的计算机方法旨在使这些扫描仪能够充分发挥其潜力，从而提高 PET 图像的准确性。拟议的工作与公共卫生相关，因为 PET 图像准确性的提高将导致更准确地诊断癌症和其他疾病、更准确地制定治疗计划以及更准确地监测治疗反应，进而导致更好的患者治疗效果。