Effect of repetition time on metabolite quantification in the human brain in 1 H MR spectroscopy at 3 tesla.

Abstract

To examine the effects of repetition time (TR) on metabolite concentration measurements in the human brain in 1 H magnetic resonance spectroscopy at 3 Tesla (T). Spectra were acquired from the posterior cingulate of five healthy adults at repetition times of 1.5 s, 3.0 s, 4.0 s, 6.0 s, and 8.0 s on a 3T MRI system. Relaxation data were also acquired for the water signal in the voxel of interest to separate tissue water and cerebrospinal fluid signal contributions. All data were quantified relative to total creatine and relative to the tissue water signal. On average, the variance for absolute metabolite concentrations was smaller than that of ratio concentrations (P = 0.003). Metabolite ratio concentrations calculated from a short TR of 1.5 s significantly differed (P < 0.05) from their "true" ratios, i.e., ratios corrected for T1 -weighting. In comparison, absolute metabolite concentrations exhibited significant differences (P < 0.05) up to a 4-s TR. To minimize potential TR-dependent concentration differences at 3T, a minimum TR of 2.5 s is suggested for ratio concentration measurements, and a 5-s TR for absolute concentrations. When possible, preference should be to perform absolute concentration measurements. 2 J. Magn. Reson. Imaging 2017;45:710-721.