A Method to Automate the Metabolite Postprocessing and Quantification on Magnetic Resonance Spectroscopy Signals

Abstract

In this paper, an automatic method to process and quantify the metabolites of a magnetic resonance spectroscopy signal is presented. Signals obtained from a region of the left thalamus were analyzed. The obtained results were compared with the ones obtained using a Phillips’ commercial magnetic resonance software. Three features of the signal corresponding to five different metabolites (NAA, Cr, Cho, Glx, Cr2) were obtained: width, height and area of the curve corresponding to each peak. The obtained results showed a maximum relative error of over 20%, indicating the potential that the proposal has. Further refinements are required to reduce the error.