Metabolite concentration ratios from single-voxel MR spectroscopyα
| Ratio | STEAM | PRESS | ||||
|---|---|---|---|---|---|---|
| Left PAβ | Right PA | Controlγ | Left PA | Right PA | Controlγ | |
| NA/Cr | 0.81 | 0.92 | 1.11 ± 0.12 | 1.16 | 1.45 | 1.61 ± 0.17 |
| Cho/Cr | 0.23 | 0.21 | 0.17 ± 0.05 | 0.23 | 0.25 | 0.23 ± 0.03 |
| m-Ins/Cr | 0.73 | 0.43 | 0.46 ± 0.05 | ndδ | nd | nd |
| Cho/NA | 0.28 | 0.23 | 0.15 ± 0.04 | 0.2 | 0.18 | 0.15 ± 0.02 |
| m-Ins/NA | 0.9 | 0.47 | 0.42 ± 0.04 | nd | nd | nd |
α Metabolite concentration ratios must be multiplied by the ratios of protons contributing to the corresponding resonances in order to obtain “peak area ratios.” The ratios for the protons are as follows: NA/Cr = 3/3, Cho/Cr = 9/3, m-Ins/Cr = ∼3.4/3, Cho/NA = 9/3, m-Ins/NA = ∼3.4/3. Thus, the “peak area ratio” for STEAM, left PA Cho/Cr = 0.23 (9/3) = 0.69. The peak area ratios calculated in this way from the concentration ratios yield results that are qualitatively in agreement with the relative peak areas in Figure 2. If the linewidths of the resonances are not excessively broadened or variable, similar qualitative agreement is achieved when relative peak amplitudes, rather than peak areas, are assessed. {Addendum: The estimate of 3.4 protons contributing to the 3.56-ppm myo-Ins resonance was based on two approaches: (1) analysis of synthetic spectra generated with the TE/TM values used in this case, (2) analysis of metabolite peak area ratios (from a standard peak fitting algorithm) relative to concentration ratios (from LCModel) in a database of 124 similarly acquired control spectra (JB Murdoch, unpublished results).}
β PA = periatrial region.
γ Control data reported as mean ± standard deviation.
δ nd = not determined.