Meeting Banner
Abstract #1780

Quantification Differences of 1H Spectra in Human Brain at 3 Tesla Using the Acquired Macromolecule Baseline or the Built-In LCModel Spline Baseline

Benoit Schaller1, Lijing Xin2, Rolf Gruetter1, 3

1Laboratory of Functional and Metabolic Imaging, Ecole Polytechnique Federale de Lausanne, Lausanne, Vaud, Switzerland; 2Department of Radiology, University of Lausanne, Lausanne, Switzerland; 3Department of Radiology, Univerisities of Lausanne and Geneva, Switzerland

Macromolecules exhibit a broad signal underlying the entire 1H spectrum and inaccuracy in the measurement of the macromolecules might lead to systematic errors preventing an accurate and reliable quantification of the metabolites. Macromolecule signal was acquired with an IR SPECIAL sequence and then inserted into the LCModel basis sets. The choice of the built-in LCModel spline baseline and acquired in vivo macromolecules (n=3) yielded either minor (Ins, Cr, PCr, Glu around 8-25%) or major (GABA, PE, GPC around 36-68%) differences in the metabolites quantification. These substantial differences are mainly due to an underestimation of the macromolecule signal around 2-3.5ppm.

Abstract PDF

Keywords

absence according accuracy acquisition adjusted aforementioned aged agreement although amplitude approved approximation assigned averaging basis blue brain broad built carefully challenged characterized choice coil committee concentrated concentration conclude contamination corrected correction cortex crucial curves deviation directly employing estimation ethics experiment field fields fixing foundations frequency frontal full functions good grant great green healthy heterogeneity highly human identified illustrates importance in vivo inaccuracy included individual influence integrated internal inversion inverted investigate judged laboratory lipid lobe local localized location macromolecule macromolecules mathematically measured metabolic metabolism metabolite metabolites middle minimize model nulled optimum outer overall peak peaks placement poor positive preceded precise preliminary previous previously prior procedure providing pulse quality quantification quantified radiology receive related relatively remain represented residual residuals resolution revealed scanner sensitivity sets shadow shape shim shims short simulated since slightly smallest special spectra spectral spectrum spin spline splines statistical subjects substantial supported systematic transmit trio typical underestimation understand vapor vector volume water years