一种新的分辨磁共振质子谱重叠弱信号的方法  

A New Method of Distinguishing Weak and Overlapping Signals of Proton Magnetic Resonance Spectroscopy

作　　者：蒋刚[1] 全红[1] 王成[1] 龚启勇[2] 

机构地区：[1]武汉大学物理科学与技术学院,武汉430072 [2]四川大学华西医院,华西磁共振研究中心,成都610041

出　　处：《生物医学工程学杂志》2012年第6期1052-1057,共6页Journal of Biomedical Engineering

基　　金：国家自然科学基金资助项目(10875092)

摘　　要：本文首次提出采用平移不变量(TI)法结合小波阈值(WT)法分辨质子磁共振波谱(1H-MRS)中的重叠弱信号。首先将1H-MRS频谱信号转化到小波域,得到谱信号的小波系数分布;然后通过TI法与WT法,得到被掩盖在大信号下的小信号。通过对模拟数据的分析表明:该算法能有效准确地获取小信号频率和幅值信息,并结合信号拟合,实现其峰下面积定量计算。In this paper, a new method of combining translation invariant （TI） and waveletthreshold （WT） algo rithm to distinguish weak and overlapping signals of proton magnetic resonance spectroscopy （1HMRS） is presen ted. First, the 1HMRS spectrum signal is transformed into wavelet domain and then its wavelet coefficients are ob tained. Then, the TI method and WT method are applied to detect the weak signals overlapped by the strong ones. Through the analysis of the simulation data, we can see that both frequency and amplitude information of smallsig nals can be obtained accurately by the algorithm, and through the combination with the method of signal fitting, quantitative calculation of the area under weak signals peaks can be realized.

关 键 词：磁共振波谱 平移不变量 小波阈值 信号拟合 定量计算 

分 类 号：R318.6[医药卫生—生物医学工程]