. Author manuscript; available in PMC: 2021 Jul 1.

Published in final edited form as: J Magn Reson. 2020 Jun 1;316:106756. doi: 10.1016/j.jmr.2020.106756

Table 2.

Comparison of metabolite quantification precisions between the multi-edit method and the proposed single-step spectral editing method by Monte Carlo analysis. Simulated MRS data were added with random noise and then quantified by spectral fitting. This simulation was repeated 300 times to compute the noise-originated CVs of the metabolite concentrations and F-tests were performed to evaluate if the two methods gave significantly different variances in metabolite concentration values.

	CV (%) multi-edit	CV (%) single-step edit	Significantly different (p < 0.01)
NAA (CH₃)	0.30	0.71	Yes
NAA (CH₂)	1.02	1.28	Yes
GABA	7.21	19.84	Yes
Glu	0.95	0.76	Yes
Gln	3.67	2.53	Yes
GSH	3.25	2.39	Yes
tCr	0.37	0.54	Yes
GPC + PC	0.43	0.48	No
mI	0.82	0.90	No

Abbreviations: CV: coefficient of variation; NAA: N-acetyl-aspartate; GABA: γ-aminobutyric acid; Glu: glutamate; Gln: glutamine; GSH: glutathione; tCr: total creatine; GPC: glycerophosphocholine; PC: phosphocholine; mI: myo-inositol.