VOLUMETRIC MR MOLECULAR IMAGING OF BRAIN

描述(由申请人提供)： 该项目将开发神经成像方法，通过短时质子磁共振光谱成像(MRSI)获取和绘制大脑代谢物的图谱。这种测量通常因光谱重叠、线形扭曲、脂质污染和受试者运动而变得复杂，这些影响限制了代谢物测量的准确性以及可能检测到的代谢物的数量。这些限制将通过使用高速采集方法来获得多个体积MRSI数据集来解决，每个数据集都用不同的光谱信息进行编码。将使用多维参数建模分析方法开发数据分析方法，以改进对重叠光谱分量的区分。此外，将同时包括更高的空间分辨率采集、场不均匀校正和运动补偿方法，以获得具有更好的空间响应和光谱质量的数据。将实施利用自旋松弛和演化参数的获取序列，并在正常受试者中进行评估。参数将使用数值模拟技术进行优化，以最大限度地提高光谱分辨率和信噪比，同时限制所需的独立光谱测量数量。这些同样的数值方法也将用于提供光谱分析程序所需的先验信息。所开发的方法将用于绘制使用短TE质子磁共振波谱在正常人脑中观察到的所有代谢物，并将生成代谢物水平的3D图像数据库。这将利用通过IRPG与旧金山MRS股(PIG)的研究人员互动获得的空间转换和归一化程序。Matson)。这一进展将建立正常代谢物图谱的范围和变异，并为使用Short-TE 1HMRS的众多临床研究提供有价值的比较数据。

DESCRIPTION (provided by applicant): This project will develop neuroimaging methods for acquisition and mapping of brain metabolites by short-TE proton MR Spectroscopic Imaging (MRSI). This measurement is typically complicated by spectral overlap, lineshape distortions, lipid contamination, and subject motion, and these effects limit the accuracy of the metabolite measurement as well as the number of metabolites that can potentially be detected. These limitations will be addressed by using high-speed acquisition methods to obtain multiple volumetric MRSI data sets that are each encoded with varying spectral information. Data analysis methods will be developed using multidimensional parametric modeling analysis approaches, to provide improved discrimination of overlapping spectral components. In addition, higher spatial resolution acquisition, field inhomogeneity correction, and motion compensation methods will be simultaneously included to obtain data with improved spatial response and spectral quality. Acquisition sequences will be implemented that exploit both spin relaxation and evolution parameters, and evaluated in normal subjects. Parameters will be optimized using numerical simulation techniques to maximize spectrat discrimination and signal to noise ratios, while also limiting the number of independent spectral measurements required. These same numerical methods will also be used to provide the prior information required for the spectral analysis procedures. The developed methods will be applied to mapping all metabolites observed using short-TE proton MR spectroscopy in normal human brain, and a 3D image database of metabolite levels will be generated. This will make use of spatial transformation and normalization procedures obtained through IRPG interactions with researchers at the MRS Unit, San Francisco (PIG. Matson). This development will establish the range and variance of normal metabolite profiles and provide valuable comparative data for numerous clinical investigations that use short-TE 1H MRS.