GENERALIZED EFFICIENT NMR SIGNAL LOCALIZATION

点击翻译按钮获取中文摘要

DESCRIPTION (Adapted from Applicant's Abstract): The broad, long-term objectives of this project are to extend and perfect a general solution to the problem of localizing nuclear magnetic resonance signals using a-priori information. This information may be combined mathematically with other signals to give the spectra, concentrations, and properties of metabolites and other substances in all anatomical regions simultaneously, but in a much shorter time than required by conventional imaging, and with fewer and less stringent demands on the imaging system than those made by less general methods. Only modest additions to signal processing software are required. The investigators have developed methodologies for these ideas, including a compartmental modeling approach (the SLIM model) and a generalized series approach (GSLIM) for compartmental imaging. In extending this concept to dynamic imaging, they have developed two novel model-based methods called RIGR (Reduced-Encoding Imaging by Generalized Series Reconstruction) and its variant TRIGR (Two-Referenced RIGR) and DIME (Dynamic Imaging by Motion Estimation). The goal of this proposal is refinement of these techniques to the point where wide-spread in-vivo applications are possible. During this continuation the applicants expect to achieve the following specific aims: (1) to improve the software tools developed during the first funding period so that they are convenient and user friendly; (2) to continue to test, by simulation and phantom experiments, theoretical predictions concerning corrections for B1 inhomogeneities, resonance offsets and other experimental nonidealities; (3) to develop criteria for prospective evaluation of the most effective data acquisition conditions for a specific study, and the probable reliability of the data which will be obtained, and criteria so that retrospective evaluation of data reliability (random and systematic errors) can be carried out; (4) to develop schemes using navigator signals and (k,t)-space sampling strategies to determine theoretically and practically the extent to which motion compensation can be accomplished in SLIM/GSLIM, RIGR/TRIGR and DIME; (5) to develop schemes using RIGR and related methods to further define possible boundaries within compartments so as to enhance the localization of SLIM and GSLIM analysis; and (6) to demonstrate the use of SLIM and GSLIM for quantitative analysis of 1H and 31P spectra of animal tissues, including muscle, brain and liver, and to compare the results with those of existing techniques.