Faster Dynamic MRI with Sparse Sampling

描述（申请人提供）：MRI在各种疾病和生理过程的动态成像方面具有巨大的潜力，但由于现有技术的成像速度有限，尚未充分用于临床应用。自MRI发明以来，对更高成像速度的追求一直是其研究的主要推动力。尽管在过去的三十年里，快速核磁共振技术取得了巨大的进步，但实际上所有的核磁共振应用都可以从额外的加速中受益，而且许多潜在的应用只有在显著加速的情况下才有可能实现。提出的项目的主要目标是通过利用稀疏采样理论的最新突破和PI小组在该领域的新工作，产生显着更快的MRI技术。这一目标将通过具体的研究工作来实现：a)开发和优化一种利用部分可分性和空间光谱约束从高度不足采样（k, t）空间数据中重建图像的新方法，b)分析和表征所提出方法的分辨率和噪声特性，以及c)评估和验证所提出的方法用于心脏成像应用，使用幻影和大鼠研究。这项研究工作的结果将在几个方面具有重要意义。首先，它将提供一个新的数学和算法框架，有效地利用多维MRI信号的稀疏性和部分可分性；该框架将实现稀疏数据采样，并显著加快当前的MRI方法。其次，它将产生新的核磁共振成像技术，提高现有核磁共振系统的性能，并为优化当前和下一代核磁共振系统的核磁共振数据采集和图像重建设计提供新的途径。第三，它将实现一系列具有挑战性的动态成像实验，包括实时3D心脏成像应用（例如，移植心脏的功能评估）。

DESCRIPTION (provided by applicant): MRI has enormous potential for dynamic imaging of various diseases and physiological processes, which has not been fully utilized for clinical applications due to the limited imaging speed of existing technology. The quest for higher imaging speeds has been a major driving force for MRI research since its invention. Although tremendous progress has been made in fast MRI technology over the last three decades, virtually all MRI applications could benefit from additional speedups, and many potential applications would become possible only with significant acceleration. The primary objective of the proposed project is to produce significantly faster MRI technology by leveraging the recent breakthroughs in sparse sampling theory and the novel work of the PI s group in this area. This objective will be achieved with specific research efforts on: a) developing and optimizing a novel method for image reconstruction from highly undersampled (k, t)-space data using both partial separability and spatial-spectral constraints, b) analyzing and characterizing the resolution and noise properties of the proposed methods, and c) evaluating and validating the proposed method for cardiac imaging applications using phantom and rat studies. The outcome of the research effort will be significant in several ways. First, it will provide a new mathematical and algorithmic framework that effectively exploits the sparsity and partial separability of multidimensional MRI signals; this framework will enable sparse data sampling and significantly accelerate current MRI methods. Second, it will produce new MR imaging technology that will enhance the performance of existing MRI systems and provide a new way to optimize the design of MR data acquisition and image reconstruction in current and next-generation MRI systems. Third, it will enable a range of challenging dynamic imaging experiments, including realtime 3D cardiac imaging applications (e.g., functional assessment of transplanted hearts). PUBLIC HEALTH RELEVANCE: This project will generate novel technology for fast dynamic magnetic resonance imaging (MRI), which will significantly enhance the clinical utility of MRI for the detection and diagnosis of diseases.