COMPRESSED SENSING APPLICATIONS TO METABOLIC IMAGING

这个子项目是许多研究子项目中的一个 由NIH/NCRR资助的中心赠款提供的资源。子项目和 研究者（PI）可能从另一个NIH来源获得了主要资金， 因此可以在其他CRISP条目中表示。所列机构为 研究中心，而研究中心不一定是研究者的研究机构。 众所周知，健康和患病组织中代谢物水平的知识可以提供关于疾病进展和治疗功效的重要信息。例如，与良性肿瘤相比，转移性肿瘤具有增加的代谢物水平，如胆碱和乳酸。三维化学位移成像（3D-CSI）是一种基于MR的非侵入性方法，用于临床定量和监测这些代谢物。3D-CSI的常规临床使用中的主要障碍是较长的采集时间，因此患者在扫描仪中花费的时间较长。因此，迫切需要解决这个问题，以使临床医生能够常规使用这种强大的技术。该项目旨在通过使用压缩传感来克服这一限制，这在过去几年中已经成为一项革命性的发明。该技术已成功应用于MRI，有望成为减少MRI扫描采集时间的新途径。我们计划对脑和乳腺CSI数据集进行回顾性分析，以便将使用传统k空间重建方法获得的代谢物图谱与使用欠采样数据进行的基于压缩感知的重建进行比较。我们假设，通过利用k空间的稀疏性以及频谱数据，我们可能能够将患者的CSI扫描时间减少2倍，而不会显着降低数据质量。

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. It is well known that knowledge of metabolite levels in healthy and diseased tissues can provide essential information on disease progression and efficacy of therapy. For instance, metastatic tumors have increased levels of metabolites like choline and lactate compared to benign tumors. 3 dimensional Chemical Shift Imaging (3D-CSI) is an MR-based non-invasive approach used in the clinic to quantify and monitor these metabolites. A major hurdle in routine clinical use of 3D-CSI is the long acquisition time and hence the time spent by the patient in the scanner. Therefore a strong need to address this problem exists to enable clinicians to make routine use of this powerful technology. The proposed project aims to overcome this limitation by the use of compressive sensing, which has been a revolutionary invention in the past few years. This technique has been successfully implemented for MRI and promises to be a new path for reducing acquisition times for MRI scans. We plan to conduct a retrospective analysis of brain and breast CSI data sets in order to compare metabolite maps obtained with conventional k-space reconstruction method to compressive sensing based reconstruction using undersampled data. We hypothesize that by exploiting the sparsity of k-space as well as the spectral data, we may be able to reduce CSI scan times for patients by a factor of 2 without significant reduction in the quality of data.