Tangi Roussel1, Hlne Ratiney1, Sophie Cavassila1
1Creatis-LRMN, CNRS UMR 5220, Inserm U630, INSA-Lyon, Universit Lyon 1, Villeurbanne, France
Many in vivo proton Magnetic Resonance Spectroscopy (MRS) studies led to one dimensional (1D) short echo time spectroscopic signal quantification methods. However, standard in vivo MRS acquisitions in 1D spectroscopy present important limitations. 1D MRS spectra usually present rich in vivo metabolic information through complicated, overlapped spectral signatures with large linewidths, especially at moderate field. Accurate concentration quantification remains problematic especially for coupled metabolites such as -aminobutyric acid (GABA), Glutamine (Gln), Glutamate (Glu), myo-inositol (mI) and Taurine (Tau). To overcome these limitations, the two dimensional (2D) spectroscopy, has great potential to unravel of the spectral information. Very few studies on 2D spectroscopic signal quantification were performed. This paper introduces a 2D spectroscopic signal quantification algorithm based on a global fitting procedure using strong prior-knowledge in the time domain and investigates the correlation between parameter estimates occurring in 1D vs 2D quantification.