A fundamentally improved MRI methodology applied to fMRI

从根本上改进的 MRI 方法应用于 fMRI

• 批准号: 6953966
• 负责人: DONALD B TWIEG
• 金额: $ 42.4万
• 依托单位: UNIVERSITY OF ALABAMA AT BIRMINGHAM
• 依托单位国家: 美国
• 财政年份: 2003
• 资助国家: 美国
• 起止时间: 2003-08-01 至 2008-06-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/6953966
• 关键词: Macaca mulatta; bioimaging /biomedical imaging; computer program /software; computer system design /evaluation; functional magnetic resonance imaging; magnetic resonance imaging; method development; phantom model; technology /technique development

DESCRIPTION (provided by applicant): Contemporary magnetic resonance imaging (MRI) methodology involves an implicit assumption, which is computationally convenient but physically inaccurate. This R21/R33 project seeks to develop a novel MRI methodology which abandons this assumption, thereby avoiding defects which commonly afflict rapid single-shot MR images. Because it interprets MRI raw data more accurately, the new methodology promises also to permit fundamentally more efficient, accurate and robust measurements to be made in several general types of MRI applications, such as diffusion imaging, measurement of tissue relaxation parameters, flow and motion imaging, and blood perfusion imaging. A strength of the general approach is that it measures tissue parameters more directly than do established approaches, which infer parameters from multiple separately acquired images. In an initial investigation and development of this methodology (termed PARSE), this project will implement a single-shot PARSE method, SS-PARSE, which is especially well-suited to functional MRI (fMRI) of the brain, with several theoretical advantages over existing methods used in fMRI. These theoretical advantages in accuracy, acquisition speed, and robustness, are substantiated in initial simulation trials. The immediate goal of the R21 project will be to experimentally verify these expected performance advantages of the SS-PARSE technique in well-controlled imaging studies using phantom objects in a 4.7T vertical-bore primate MR imaging system. A shortcoming of the methodology is that it requires much longer computation times than conventional MRI approaches. The R21 project will seek to develop much faster computational algorithms, and will determine the prospects for reducing computational time to roughly that of typical conventional fMRI studies. If these two major performance milestones have been met, the R33 project will commence, testing and characterizing the new methodology in functional brain imaging studies of primates. Additionally, the R33 project will include development of more sophisticated versions of the SS-PARSE method for high-resolution functional imaging. The project promises to introduce a new methodology, which may have substantial fundamental performance benefits over a broad range of clinical MRI applications.

描述（由申请人提供）： 当代磁共振成像（MRI）方法涉及隐含假设，该假设在计算上方便但在物理上不准确。 该 R21/R33 项目旨在开发一种新颖的 MRI 方法，该方法放弃了这一假设，从而避免了通常影响快速单次 MR 图像的缺陷。 由于新方法能够更准确地解释 MRI 原始数据，因此还有望在几种常见类型的 MRI 应用中进行更高效、更准确和更稳健的测量，例如扩散成像、组织松弛参数测量、血流和运动成像和血液灌注成像。 一般方法的优点是它比现有方法更直接地测量组织参数，现有方法从多个单独采集的图像推断参数。 在对该方法（称为 PARSE）的初步研究和开发中，该项目将实施单次 PARSE 方法 SS-PARSE，该方法特别适合大脑的功能 MRI (fMRI)，与功能磁共振成像中使用的现有方法。 这些在准确性、采集速度和鲁棒性方面的理论优势在初始模拟试验中得到了证实。 R21 项目的近期目标是通过实验验证 SS-PARSE 技术在 4.7T 垂直孔灵长类 MR 成像系统中使用模型物体的良好控制成像研究中的这些预期性能优势。 该方法的一个缺点是它需要比传统 MRI 方法更长的计算时间。 R21 项目将寻求开发更快的计算算法，并将确定将计算时间减少到与典型的传统 fMRI 研究大致相同的前景。 如果这两个主要的性能里程碑得以实现，R33 项目将开始，测试和表征灵长类动物功能性脑成像研究的新方法。 此外，R33 项目将包括开发用于高分辨率功能成像的 SS-PARSE 方法的更复杂版本。 该项目有望引入一种新的方法，该方法可能对广泛的临床 MRI 应用具有实质性的基本性能优势。