Assessment of measurement precision in single‐voxel spectroscopy at 7 T: Toward minimal detectable changes of metabolite concentrations in the human brain in vivo

Abstract

This is an open access article under the terms of the Creat ive Commo ns Attri bution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Fourier‐based decomposition for simultaneous 2‐voxel MRS acquisition with 2SPECIAL

“…The SB WURST pulse was already successfully utilized in other MR spectroscopy sequences 14,40,41 but has – to our knowledge – not been used as an MB pulse before, neither in MR imaging nor in MRS. For SVS acquisitions, it was recently shown that the WURST pulse does not exhibit significant differences in the reproducibility of resulting metabolite concentrations and spectral shape analysis but allows for decreased peak power and reduced chemical shift displacement, compared to the HS pulse that is used by default in the SPECIAL sequence 24 . This is even more important for MB pulses because, in the simplest case, nearly twice the nominal peak power is needed to fulfill the adiabatic condition 20 .…”

“…The sequence diagram for the 2SPECIAL sequence employing a WURST inversion pulse before the excitation is depicted in Figure 1A. Bloch simulations were performed taking the transmit magnetic field B 1 + as well as the off‐resonances ΔB 0 into account to evaluate the excitation profile performance and compare it to the MB HS pulse 23,24 …”

“…Boxplots of the corrected concentration $$$ {c}_{i,m,v}^{\ast } $$$ and the Cramér‐Rao lower bounds (CRLBs) over all subjects $$$ i $$$ and each voxel $$$ v $$$ for every metabolite $$$ m $$$ were derived to evaluate the differences between (1) SVS ind and SVS comb , (2) SVS ind and both decomposed sMVS spectra, (3) SVS comb and both decomposed sMVS spectra, and (4) vGRAPPA and SENSE‐based decomposed spectra over all ten subjects. Moreover, Bland–Altman plots 38 reflecting differences of the real part of the spectral shape for (3) were obtained to assess differences in the spectra without any bias by the fit required for metabolite quantification 24 …”

“…Bloch simulations were performed taking the transmit magnetic field B 1 + as well as the off-resonances ΔB 0 into account to evaluate the excitation profile performance and compare it to the MB HS pulse. 23,24…”

“…Moreover, Bland-Altman plots 38 reflecting differences of the real part of the spectral shape for (3) were obtained to assess differences in the spectra without any bias by the fit required for metabolite quantification. 24 A nonparametric paired Wilcoxon signed-rank test 39 was performed to investigate statistical differences of the metabolite concentrations and CRLBs between the following pairs of methods: (1) SVS comb and SVS ind , (2) SENSE-based decomposition and SVS comb , (3) vGRAPPA decomposition and SVS comb , (4) SENSE-based decomposition and SVS ind , (5) vGRAPPA decomposition and SVS ind , and (6) vGRAPPA and SENSE-based decomposition. Due to Bonferroni correction, the significance level of P < 0.05 was shifted to P < 0.0038 because the 13 metabolites cannot be fitted independently of each other.…”