Quantification of human brain metabolites from in vivo1H NMR magnitude spectra using self-organising maps

Abstract

The self-organising map (SOM) analysis has been successfully applied in many fields of research and it is a potential tool also for analysis of magnetic resonance spectroscopy (MRS) data. In this paper we demonstrate that SOM-based analysis, can be applied for automated MRS data quantification. To this end, a set of experimental long echo time (TE=270 ms) in vivo 1H MRS spectra were initially analysed by the lineshape fitting (LF) method to find out simulated spectra mathing to the experimental data. The results from simulated data sets show that clinically relevant metabolite quantification from human brain MRS can be obtained with the SOM analysis. © Springer-Verlag 2003.