Separation of EMCL and IMCL in Musculoskeletal 1H MR Spectra by Filter-Diagonalization Method (FDM)

Jyh-Miin Lin¹, Shang-Yueh Tsai²

Time domain algorithm is useful to quantify the muscular 1H MR spectra but is less efficient and not frequency-selective. Here we propose to use the Filter-Diagonalization Method (FDM), a frequency-selective time domain algorithm, originally developed for quantum dynamic computation, to separate the extramyocellular lipids (EMCL) and intramyocellular lipids (IMCL). The local basis can extract the metabolite signals in muscular 1H MRS data with economic computational cost. Results from our preliminary study showed, with a properly selected basis, the FDM can reliably separates the near CH2 groups of EMCL(~1.49ppm) and IMCL(~1.28ppm).