The Quantification Study Of Multi-Voxel Magnetic Resonance Spectroscopy In Magnetic Resonance Imaging

Magnetic resonance spectroscopy (MRS) is result of phenomenon of nuclear magnetic resonance(NMR) and chemical shift of nucleus in the organization to be tested, which is an important supplement of the magnetic resonance imaging. MRS can quantitatively reflect changes of various types of compound in the organization. It is a kind of reflection of the living tissue metabolism, known as "virtual biopsy". MRS is a non-invasive, non-destructive testing methods of imaging, which provides an important reference for the diagnosis and treatment of disease. Due to the influence of environmental conditions and physical activity and so on, MRS can not be directly used in clinical diagnosis. Thus, MRS need to go through a series of post-processing. Than MRS can be applied to clinical diagnosis. MRS process is divided into two steps:preprocessing and quantification. Preprocessing is precisely quantified pretreatment premise.This paper focuses on multi-voxel magnetic resonance spectroscopy analysis. Because multi-voxel magnetic resonance spectroscopy contains dozens or even hundreds of spectrum and not all of the spectrum are useful, we need to choose a number of interested spectrum to process. In order to achieve this purpose, firstly,we must realize visualization of multi-voxel magnetic resonance spectroscopy. In order to achieve visualization, First of all, we need spatial location information of MR image and multi-voxel magnetic resonance spectroscopy, which can be abtained by reading the Tag number of DICOM files and the header file*.spar of Philips magnetic resonance spectroscopy. Secondly, according to the spatial location information and the relationship between the field of vision, volume of interest and voxels, the use of mathematical knowledge can achive visualization.In order to facilitate the process of multi-voxel magnetic resonance spectroscopy, the design achives a set of the visualization of multi-voxel magnetic resonance spectroscopy, preprocessing(denoising,water supressed and correction of phase etc.), quantification(using HSVD), statistical analysis of the magnitude of metabolites and voxel localization as one tool based on MATLAB GUI, which can enable averaging and individual analysis on multi-voxel magnetic resonance spectroscopy. Averaging refers to calculate the mean spectroscopy of selected several spectroscopy. Statistical analysis can analyze NAA, Cho and Cr in statistics, including box polts, histograms and the mean, standard deviation. In histograms, metabolites are arranged according to the order of selsected voxels. According to the order, the operator can quickly, accurately positioning the interested metabolites to the corresponding voxel in registration image. In this design, multi-voxel magnetic resonance spectroscopy analysis is done automatically, which effectively reduces required technology knowledge of the operator.