Automatic Data Processing to Relative Quantitative Analysis of 1H MR Spectroscopy of Brain

This paper presents an automatic method for processing the data of 1H magnetic resonance spectroscopy. This method focuses on expressing the metabolites concentration ratios of the brain, such as NAA/Cr, NAA/Cho, Cho/Cr, etc. The data processing system consists of water suppression with wavelet basis function, automatic coefficients selection of the Gaussian transform based on the signal characteristics and experience formulas. Especially, our automatically relative quantitative analysis is based on an approximate method to label the chemical shift in parts per million (ppm) in the case of some necessary information are missed, such as the sampling rate and reference resonance frequency. It is very useful in the fusion system of brain tumor features extracted from MR images and MR spectrum.