- Email: [email protected]
Cerebral Magnetic Resonance Spectroscopy at 7 Tesla
Cerebral Magnetic Resonance Spectroscopy at 7 Tesla: Standard Values and Regional Differences Astrid E. Grams, MD, Irina Brote, MSc, Stefan Maderwald, PhD, Kiriaki Kollia, MD, Mark E. Ladd, PhD, Michael Forsting, MD, Elke R. Gizewski, MD Rationale and Objectives: At lower magnetic field strengths, regional differences of cerebral metabolite distributions have been described, but these data are controversial. Magnetic resonance spectroscopy at 7 T is expected to deliver high spectral resolution and good differentiation, but there are problems arising at high magnetic field strengths that may diminish spectral quality. Because there have been only a few studies in humans so far, there are no standard values for 7 T concerning regional metabolite distributions and concentrations. Materials and Methods: In the present study, the metabolites detectable with 1H magnetic resonance spectroscopy, N-acetyl-aspartate, choline, and creatine (Cr), were evaluated with a single-voxel sequence. Five voxels were placed in the frontal and parietal white matter and the insular, thalamic, and occipital gray matter. Results: For N-acetyl-aspartate, the lowest values were found in frontal white matter and the highest in thalamic gray matter. Choline displayed the lowest values in frontal white matter and the highest in insular gray matter. Cr showed the lowest values in frontal white matter and the highest in thalamic gray matter. The highest ratio of choline to Cr was found in parietal white matter and the lowest in thalamic gray matter. The highest ratio of N-acetyl-aspartate to Cr was found in thalamic gray matter and the lowest in frontal white matter. Conclusions: In the present study, regional cerebral metabolite differences were verified with high-field magnetic resonance spectroscopy. Quantitative values and metabolite ratios could be a basis for further clinical studies. Key Words: MRI; spectroscopy; MRS; regional; 7 tesla. ªAUR, 2011
I
n prior studies at 1.5 and 3 T, regional differences of cerebral metabolite concentrations have been described, but these data are contradictory (1–4). At 7 T, cerebral regional metabolite distributions have not yet been evaluated. High-field magnetic resonance spectroscopy (MRS) has the potential to provide enhanced neurochemical information on the basis of increased sensitivity and higher spectral resolution (5). However, problems arising in highfield magnetic resonance imaging, such as more pronounced B0 and B1 inhomogeneities, may decrease spectral resolution and minimize quantitation accuracy. The present 7-T study
Acad Radiol 2011; 18:584–587 From the Department of Neuroradiology, University Hospital Giessen and Marburg, Justus Liebig University Giessen, Klinikstraße 29, 35392 Giessen, Germany (A.E.G., E.R.G.); the Department of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital Essen, University Duisburg-Essen, Essen, Germany (A.E.G., I.B., S.M., K.K., M.E.L., M.F., E.R.G.); and the Erwin L. Hahn Institute for Magnetic Resonance Imaging, University Duisburg-Essen, Essen, Germany (A.E.G., I.B., S.M., K.K., M.E.L., M.F., E.R.G.). Received August 8, 2010; accepted December 16, 2010. Address correspondence to: A.E.G. e-mail: [email protected]. uni-giessen.de ªAUR, 2011 doi:10.1016/j.acra.2010.12.010
584
was an attempt to clarify existing uncertainties concerning cerebral metabolite distributions with established preprocessing software (6,7) and quantization algorithm (8). For this purpose, the three most commonly used metabolites, N-acetyl-aspartate (NAA), choline (Cho) and creatine (Cr), were evaluated. NAA is a marker for neuronal density, integrity, and viability and resonates at a frequency of 2.02 ppm (9). Cho is a marker of membrane density and integrity; its resonance frequency is at 3.22 ppm. Cr, which resonates at 3.02 ppm, is a marker for energy metabolism and is assumed to be quite stable in healthy but also pathologic brain tissue (10).
MATERIALS AND METHODS Ten healthy volunteers aged 22 to 31 years (mean age, 26 years; five women, five men) with no morphologic or pathologic changes or histories of psychiatric disorders were examined. Written informed consent was obtained prior to the scans. Data acquisition was performed using a 7-T whole-body scanner (Magnetom 7T; Siemens Medical Systems, Erlangen, Germany) and a custom-built eightchannel transmit/receive head coil (11). The study was approved by the local ethics committee.
Academic Radiology, Vol 18, No 5, May 2011
MRS was performed using a 1H single-voxel pointresolved spectroscopic sequence (scan time, 5 minutes 6 seconds; voxel size, 10 mm; repetition time, 2,320 ms; echo time, 30 ms; flip angle, 90 ; flip angle of refocusing pulses, 180 , reduced by 25% because of limited peak radiofrequency power; averages, 128). Different brain areas were evaluated: the frontal and parietal white matter and the insular, thalamic, and occipital gray matter. Positioning of the voxels was based on a three-dimensional magnetization-prepared rapid gradient-echo sequence (Fig 1). One measurement was performed in each volunteer. The water signal was suppressed by using water suppression enhanced through T1 effects. With first-order and secondorder B0 shimming, a water line width (full width at half maximum) of approximately 18 Hz was achieved. According to the chemical shift displacement, the metabolites NAA, Cho, and Cr had 100%, 71.7%, and 75.6% of their corresponding voxels in common. Spectral preprocessing including water peak removal with a Hankel-Lanczos singular value decomposition filter, phase correction of first and second order, zero filling, apodization with a Lorentzian line shape of 10 Hz, and baseline correction were performed with the software jMRUI version 3.0 (6,7). The metabolites were quantified with the advanced method for accurate, robust, and efficient spectral fitting (8). In addition, comparisons between metabolite concentration and Cho/Cr and NAA/ Cr ratios were made for the different brain regions. For statistical evaluation, a Kruskal-Wallis (more than two-group comparisons) or a Mann-Whitney test (two-group comparisons) was applied. RESULTS The targeted metabolites could be identified in all brain regions. Examples for regional metabolite differences are shown in Figure 1 and Table 1. Single-voxel spectroscopy results indicated a high signal-to-noise ratio (SNR) except in the insular gray matter, which suffered from a noticeably lower SNR (Fig 1c). For NAA, the highest concentration was found in the thalamic cortex and the lowest in the insular cortex (Fig 2a). A significant difference could be detected between the different brain areas (P < .0001). Furthermore, NAA was found to be significantly lower in white matter than in gray matter (P = .004). Cho displayed the highest concentrations in the thalamic cortex and the lowest in the occipital cortex, with a significant difference between the brain areas (P = .02). No significant difference was found between gray and white matter (Fig 2b). Cr was found to be highest in the thalamic cortex and lowest in the insular cortex, with a significant difference between the brain areas (P < .0001). For Cr, a significantly lower concentration was found in white than in gray matter (P = .003; Fig 2c). The Cho/Cr ratio displayed the highest concentration in the insular cortex and the lowest in the occipital cortex,
7 TESLA MR SPECTROSCOPY
with a significant difference between the evaluated brain areas (P < .0001). The NAA/Cr ratio did not display significant differences between the different brain areas (Table 1). DISCUSSION In the present study, regional differences of metabolite distributions in the human brain could be confirmed on 7-T MRS. In several earlier studies, regional differences of cerebral metabolite distributions in proton-based MRS were described, but none of the studies examined as many supratentorial brain regions concerning all three metabolites as the present one. In comparisons between frontal and parietal white matter, no significant differences were found for Cho, Cr, or NAA in prior studies (3,4), whereas in the present study, higher concentrations in the parietal white matter in comparison to the frontal cortex were found for all metabolites. A comparison between the insular and occipital cortices showed higher values for NAA in the occipital cortex and for Cho in the insular cortex (3), which was confirmed in the present study. No significant differences in Cr between the two areas were found, whereas in the present study, higher values were found in the occipital cortex. In a comparative study between the insular cortex and thalamic gray matter (12), higher values of NAA were found in the insular cortex, whereas in the present study, the contrary was seen. Cr was described to be even in both areas, whereas in the present study, it was found to be higher in the thalamus; Cho was found to be higher in the thalamic cortex than in the insular cortex, whereas in the present study, no significant differences were seen. Several prior studies compared metabolite distributions between white and gray matter. The most recent 1.5-T study described a significantly lower NAA concentration in white matter in comparison to gray matter (1), as was confirmed in the present study and in other prior studies (2,13). In other studies, significantly higher NAA concentrations were found in white matter than in gray matter (4,14); in one study, no significant regional differences were found for NAA (3). Cho was determined to be higher in white than in gray matter in several prior studies (2–4,14) and lower in white matter in one prior study (13), whereas in the present study, no significant differences between gray and white matter could be described for Cho. As in the present study, Cr was found to be lower in white than in gray matter in several earlier studies (2–4,13,15). In the present study, standard deviations varied up to almost 100% for some metabolite values. This effect was most severe in the insular cortex and for the metabolite Cho. A reason for this might be tissue heterogeneity in the insular cortex, with resulting susceptibility effects and signal loss. Cho is the metabolite with the lowest concentration (in comparison to NAA and Cr), so it might be most sensitive to reduced SNR. In theory, linear gains in SNR and spectral resolution are expected with increasing field strength. These gains in SNR 585
GRAMS ET AL
Academic Radiology, Vol 18, No 5, May 2011
Figure 1. Spectral results from different brain regions: (a) frontal white matter (FWM), (b) parietal white matter (PWM), (c) insular gray matter (IGM), (d) thalamic gray matter (TGM), and (e) occipital gray matter (OGM). Cho, choline; Cr, creatine; NAA, N-acetyl-aspartate.
TABLE 1. Results in the 10 Volunteers for Metabolite Concentrations Quantified Using the Advanced Method for Accurate, Robust, and Efficient Spectral Fitting (Arbitrary Units) Region FWM PWM IGM TGM OGM
NAA
Cho
Cr
Cho/Cr
NAA/Cr
13.60 5.45 19.10 3.50 12.33 5.45 38.27 8.60 31.57 6.00
7.51 2.76 9.00 7.48 8.87 8.46 9.38 6.27 3.68 2.48
7.97 2.44 10.32 1.88 7.73 2.36 20.79 4.99 18.48 3.76
0.97 0.33 0.94 0.91 1.04 0.73 0.48 0.31 0.20 0.10
1.67 0.26 1.88 0.35 1.57 0.46 1.86 0.26 1.73 0.27
Cho, choline; Cr, creatine; FWM, frontal white matter; IGM, insular gray matter; NAA, N-acetyl-aspartate; OGM, occipital gray matter; PWM, parietal white matter; TGM, thalamic gray matter.
and spectral resolution were verified experimentally in a proton-based spectroscopic imaging study in a comparison of 1.5-T, 3-T, 4-T, and 7-T scans (16). Other studies showed average SNR increases between 23% and 46% for 3-T compared to 1.5-T field strength (17,18). In a comparison between 4-T and 7-T field strength, average SNR at 7 T was only 1.6 times of that at 4 T (19). These results indicate that the theoretical degree of spectroscopic quality improvement cannot fully be reached in vivo. The results of the present study prove this, as received signal and resulting spectral quality were only moderate. There are several possible reasons for this. In the present study, a line width of water after shimming of about 18 Hz was obtained. As shown previously, line width increases with increasing field strength (18), which 586
has a diminishing effect on SNR. Because the true flip angle of the refocusing pulses was 155 and not 180 because of limited peak radiofrequency power, a loss in SNR had to be expected. To gain more SNR, a 32-channel transmit/receive head coil, for example, could be used in future studies. Other limitations arising in high-field magnetic resonance imaging that may decrease spectral quality include B0 or B1 inhomogeneities and susceptibility effects, which are more pronounced with increasing magnetic field strength. Furthermore, in the present study, the insular cortex suffered from low SNR, presumably because of susceptibility effects from tissue heterogeneities. On the other hand, chemical shift increases with magnetic field strength, which can be used to better differentiate
Academic Radiology, Vol 18, No 5, May 2011
7 TESLA MR SPECTROSCOPY
Figure 2. Box plots visualizing regional distributions and differences of (a) N-acetyl-aspartate (NAA), (b) choline (Cho), and (c) creatine (Cr) quantified using the advanced method for accurate, robust, and efficient spectral fitting. FWM, frontal white matter; IGM, insular gray matter; OGM, occipital gray matter; PWM, parietal white matter; TGM, thalamic gray matter.
metabolites. With 7-T MRS, it has been shown that lowconcentration metabolites such as glycine can be detected (20), and contiguous metabolites such as Cr and phosphocreatine can be reliably differentiated from one another (5). The evaluation of the distribution of these low-concentration metabolites could be a focus for further applications. In addition, future plans include the optimization of the pointresolved spectroscopic sequence to obtain shorter repetition and echo times to enable high-efficiency measurement of J-modulating compounds (glutamate, glutamine) and short– echo time compounds such as macromolecules. CONCLUSIONS Regional differences between cerebral metabolite distributions were confirmed. These data can therefore provide a basis for further clinical applications of proton-based cerebral MRS in diseases that alter cerebral metabolites, including ischemia, tumors, and inflammatory or psychiatric disorders. REFERENCES 1. Inglese M, Rusinek H, George IC, et al. Global average gray and white matter N-acetylaspartate concentration in the human brain. Neuroimage 2008; 41:270–276. 2. Noworolski SM, Nelson SJ, Henry RG, et al. High spatial resolution 1H-MRSI and segmented MRI of cortical gray matter and subcortical white matter in three regions of the human brain. Magn Reson Med 1999; 41: 21–29. 3. Pouwels PJ, Frahm J. Regional metabolite concentrations in human brain as determined by quantitative localized proton MRS. Magn Reson Med 1998; 39:53–60. 4. Schuff N, Ezekiel F, Gamst AC, et al. Region and tissue differences of metabolites in normally aged brain using multislice 1H magnetic resonance spectroscopic imaging. Magn Reson Med 2001; 45:899–907. 5. Tkac I, Andersen P, Adriany G, et al. In vivo 1H NMR spectroscopy of the human brain at 7 T. Magn Reson Med 2001; 46:451–456.
6. Mazzoni LN, Belli G, Ginestroni A, et al. Computation of brain metabolite ratios in single-voxel proton MR spectroscopy: comparison between semiautomatic and automatic software. Radiol Med 2010; 115: 125–132. 7. Torriani M, Thomas BJ, Halpern EF, et al. Intramyocellular lipid quantification: repeatability with 1H MR spectroscopy. Radiology 2005; 236: 609–614. 8. Vanhamme L, van den Boogaart A, Van Huffel S. Improved method for accurate and efficient quantification of MRS data with use of prior knowledge. J Magn Reson 1997; 129:35–43. 9. Arnold DL, de Stefano N, Matthews PM, et al. N-acetylaspartate: usefulness as an indicator of viable neuronal tissue. Ann Neurol 2001; 50:823–825. 10. Soares DP, Law M. Magnetic resonance spectroscopy of the brain: review of metabolites and clinical applications. Clin Radiol 2009; 64:12–21. €fer LC, et al. 8-channel transmit/receive head coil 11. Orzada S, Kraff O, Scha for 7 T human imaging using intrinsically decoupled strip line elements with meanders. Presented at: ISMRM 17th Scientific Meeting & Exhibition; Honolulu, HI; 2009. 12. Frahm J, Bruhn H, Gyngell ML, et al. Localized proton NMR spectroscopy in different regions of the human brain in vivo. Relaxation times and concentrations of cerebral metabolites. Magn Reson Med 1989; 11:47–63. 13. Wang Y, Li SJ. Differentiation of metabolic concentrations between gray matter and white matter of human brain by in vivo 1H magnetic resonance spectroscopy. Magn Reson Med 1998; 39:28–33. 14. Wiedermann D, Schuff N, Matson GB, et al. Short echo time multislice proton magnetic resonance spectroscopic imaging in human brain: metabolite distributions and reliability. Magn Reson Imaging 2001; 19: 1073–1080. 15. Hetherington HP, Mason GF, Pan JW, et al. Evaluation of cerebral gray and white matter metabolite differences by spectroscopic imaging at 4.1T. Magn Reson Med 1994; 32:565–571. 16. Otazo R, Mueller B, Ugurbil K, et al. Signal-to-noise ratio and spectral linewidth improvements between 1.5 and 7 Tesla in proton echo-planar spectroscopic imaging. Magn Reson Med 2006; 56:1200–1210. 17. Barker PB, Hearshen DO, Boska MD. Single-voxel proton MRS of the human brain at 1.5T and 3.0T. Magn Reson Med 2001; 45:765–769. 18. Gonen O, Gruber S, Li BS, et al. Multivoxel 3D proton spectroscopy in the brain at 1.5 versus 3.0 T: signal-to-noise ratio and resolution comparison. AJNR Am J Neuroradiol 2001; 22:1727–1731. 19. Vaughan JT, Garwood M, Collins CM, et al. 7T vs. 4T: RF power, homogeneity, and signal-to-noise comparison in head images. Magn Reson Med 2001; 46:24–30. 20. Gambarota G, Mekle R, Xin L, et al. In vivo measurement of glycine with short echo-time 1H MRS in human brain at 7 T. Magma 2009; 22:1–4.
587
I
n prior studies at 1.5 and 3 T, regional differences of cerebral metabolite concentrations have been described, but these data are contradictory (1–4). At 7 T, cerebral regional metabolite distributions have not yet been evaluated. High-field magnetic resonance spectroscopy (MRS) has the potential to provide enhanced neurochemical information on the basis of increased sensitivity and higher spectral resolution (5). However, problems arising in highfield magnetic resonance imaging, such as more pronounced B0 and B1 inhomogeneities, may decrease spectral resolution and minimize quantitation accuracy. The present 7-T study
Acad Radiol 2011; 18:584–587 From the Department of Neuroradiology, University Hospital Giessen and Marburg, Justus Liebig University Giessen, Klinikstraße 29, 35392 Giessen, Germany (A.E.G., E.R.G.); the Department of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital Essen, University Duisburg-Essen, Essen, Germany (A.E.G., I.B., S.M., K.K., M.E.L., M.F., E.R.G.); and the Erwin L. Hahn Institute for Magnetic Resonance Imaging, University Duisburg-Essen, Essen, Germany (A.E.G., I.B., S.M., K.K., M.E.L., M.F., E.R.G.). Received August 8, 2010; accepted December 16, 2010. Address correspondence to: A.E.G. e-mail: [email protected]. uni-giessen.de ªAUR, 2011 doi:10.1016/j.acra.2010.12.010
584
was an attempt to clarify existing uncertainties concerning cerebral metabolite distributions with established preprocessing software (6,7) and quantization algorithm (8). For this purpose, the three most commonly used metabolites, N-acetyl-aspartate (NAA), choline (Cho) and creatine (Cr), were evaluated. NAA is a marker for neuronal density, integrity, and viability and resonates at a frequency of 2.02 ppm (9). Cho is a marker of membrane density and integrity; its resonance frequency is at 3.22 ppm. Cr, which resonates at 3.02 ppm, is a marker for energy metabolism and is assumed to be quite stable in healthy but also pathologic brain tissue (10).
MATERIALS AND METHODS Ten healthy volunteers aged 22 to 31 years (mean age, 26 years; five women, five men) with no morphologic or pathologic changes or histories of psychiatric disorders were examined. Written informed consent was obtained prior to the scans. Data acquisition was performed using a 7-T whole-body scanner (Magnetom 7T; Siemens Medical Systems, Erlangen, Germany) and a custom-built eightchannel transmit/receive head coil (11). The study was approved by the local ethics committee.
Academic Radiology, Vol 18, No 5, May 2011
MRS was performed using a 1H single-voxel pointresolved spectroscopic sequence (scan time, 5 minutes 6 seconds; voxel size, 10 mm; repetition time, 2,320 ms; echo time, 30 ms; flip angle, 90 ; flip angle of refocusing pulses, 180 , reduced by 25% because of limited peak radiofrequency power; averages, 128). Different brain areas were evaluated: the frontal and parietal white matter and the insular, thalamic, and occipital gray matter. Positioning of the voxels was based on a three-dimensional magnetization-prepared rapid gradient-echo sequence (Fig 1). One measurement was performed in each volunteer. The water signal was suppressed by using water suppression enhanced through T1 effects. With first-order and secondorder B0 shimming, a water line width (full width at half maximum) of approximately 18 Hz was achieved. According to the chemical shift displacement, the metabolites NAA, Cho, and Cr had 100%, 71.7%, and 75.6% of their corresponding voxels in common. Spectral preprocessing including water peak removal with a Hankel-Lanczos singular value decomposition filter, phase correction of first and second order, zero filling, apodization with a Lorentzian line shape of 10 Hz, and baseline correction were performed with the software jMRUI version 3.0 (6,7). The metabolites were quantified with the advanced method for accurate, robust, and efficient spectral fitting (8). In addition, comparisons between metabolite concentration and Cho/Cr and NAA/ Cr ratios were made for the different brain regions. For statistical evaluation, a Kruskal-Wallis (more than two-group comparisons) or a Mann-Whitney test (two-group comparisons) was applied. RESULTS The targeted metabolites could be identified in all brain regions. Examples for regional metabolite differences are shown in Figure 1 and Table 1. Single-voxel spectroscopy results indicated a high signal-to-noise ratio (SNR) except in the insular gray matter, which suffered from a noticeably lower SNR (Fig 1c). For NAA, the highest concentration was found in the thalamic cortex and the lowest in the insular cortex (Fig 2a). A significant difference could be detected between the different brain areas (P < .0001). Furthermore, NAA was found to be significantly lower in white matter than in gray matter (P = .004). Cho displayed the highest concentrations in the thalamic cortex and the lowest in the occipital cortex, with a significant difference between the brain areas (P = .02). No significant difference was found between gray and white matter (Fig 2b). Cr was found to be highest in the thalamic cortex and lowest in the insular cortex, with a significant difference between the brain areas (P < .0001). For Cr, a significantly lower concentration was found in white than in gray matter (P = .003; Fig 2c). The Cho/Cr ratio displayed the highest concentration in the insular cortex and the lowest in the occipital cortex,
7 TESLA MR SPECTROSCOPY
with a significant difference between the evaluated brain areas (P < .0001). The NAA/Cr ratio did not display significant differences between the different brain areas (Table 1). DISCUSSION In the present study, regional differences of metabolite distributions in the human brain could be confirmed on 7-T MRS. In several earlier studies, regional differences of cerebral metabolite distributions in proton-based MRS were described, but none of the studies examined as many supratentorial brain regions concerning all three metabolites as the present one. In comparisons between frontal and parietal white matter, no significant differences were found for Cho, Cr, or NAA in prior studies (3,4), whereas in the present study, higher concentrations in the parietal white matter in comparison to the frontal cortex were found for all metabolites. A comparison between the insular and occipital cortices showed higher values for NAA in the occipital cortex and for Cho in the insular cortex (3), which was confirmed in the present study. No significant differences in Cr between the two areas were found, whereas in the present study, higher values were found in the occipital cortex. In a comparative study between the insular cortex and thalamic gray matter (12), higher values of NAA were found in the insular cortex, whereas in the present study, the contrary was seen. Cr was described to be even in both areas, whereas in the present study, it was found to be higher in the thalamus; Cho was found to be higher in the thalamic cortex than in the insular cortex, whereas in the present study, no significant differences were seen. Several prior studies compared metabolite distributions between white and gray matter. The most recent 1.5-T study described a significantly lower NAA concentration in white matter in comparison to gray matter (1), as was confirmed in the present study and in other prior studies (2,13). In other studies, significantly higher NAA concentrations were found in white matter than in gray matter (4,14); in one study, no significant regional differences were found for NAA (3). Cho was determined to be higher in white than in gray matter in several prior studies (2–4,14) and lower in white matter in one prior study (13), whereas in the present study, no significant differences between gray and white matter could be described for Cho. As in the present study, Cr was found to be lower in white than in gray matter in several earlier studies (2–4,13,15). In the present study, standard deviations varied up to almost 100% for some metabolite values. This effect was most severe in the insular cortex and for the metabolite Cho. A reason for this might be tissue heterogeneity in the insular cortex, with resulting susceptibility effects and signal loss. Cho is the metabolite with the lowest concentration (in comparison to NAA and Cr), so it might be most sensitive to reduced SNR. In theory, linear gains in SNR and spectral resolution are expected with increasing field strength. These gains in SNR 585
GRAMS ET AL
Academic Radiology, Vol 18, No 5, May 2011
Figure 1. Spectral results from different brain regions: (a) frontal white matter (FWM), (b) parietal white matter (PWM), (c) insular gray matter (IGM), (d) thalamic gray matter (TGM), and (e) occipital gray matter (OGM). Cho, choline; Cr, creatine; NAA, N-acetyl-aspartate.
TABLE 1. Results in the 10 Volunteers for Metabolite Concentrations Quantified Using the Advanced Method for Accurate, Robust, and Efficient Spectral Fitting (Arbitrary Units) Region FWM PWM IGM TGM OGM
NAA
Cho
Cr
Cho/Cr
NAA/Cr
13.60 5.45 19.10 3.50 12.33 5.45 38.27 8.60 31.57 6.00
7.51 2.76 9.00 7.48 8.87 8.46 9.38 6.27 3.68 2.48
7.97 2.44 10.32 1.88 7.73 2.36 20.79 4.99 18.48 3.76
0.97 0.33 0.94 0.91 1.04 0.73 0.48 0.31 0.20 0.10
1.67 0.26 1.88 0.35 1.57 0.46 1.86 0.26 1.73 0.27
Cho, choline; Cr, creatine; FWM, frontal white matter; IGM, insular gray matter; NAA, N-acetyl-aspartate; OGM, occipital gray matter; PWM, parietal white matter; TGM, thalamic gray matter.
and spectral resolution were verified experimentally in a proton-based spectroscopic imaging study in a comparison of 1.5-T, 3-T, 4-T, and 7-T scans (16). Other studies showed average SNR increases between 23% and 46% for 3-T compared to 1.5-T field strength (17,18). In a comparison between 4-T and 7-T field strength, average SNR at 7 T was only 1.6 times of that at 4 T (19). These results indicate that the theoretical degree of spectroscopic quality improvement cannot fully be reached in vivo. The results of the present study prove this, as received signal and resulting spectral quality were only moderate. There are several possible reasons for this. In the present study, a line width of water after shimming of about 18 Hz was obtained. As shown previously, line width increases with increasing field strength (18), which 586
has a diminishing effect on SNR. Because the true flip angle of the refocusing pulses was 155 and not 180 because of limited peak radiofrequency power, a loss in SNR had to be expected. To gain more SNR, a 32-channel transmit/receive head coil, for example, could be used in future studies. Other limitations arising in high-field magnetic resonance imaging that may decrease spectral quality include B0 or B1 inhomogeneities and susceptibility effects, which are more pronounced with increasing magnetic field strength. Furthermore, in the present study, the insular cortex suffered from low SNR, presumably because of susceptibility effects from tissue heterogeneities. On the other hand, chemical shift increases with magnetic field strength, which can be used to better differentiate
Academic Radiology, Vol 18, No 5, May 2011
7 TESLA MR SPECTROSCOPY
Figure 2. Box plots visualizing regional distributions and differences of (a) N-acetyl-aspartate (NAA), (b) choline (Cho), and (c) creatine (Cr) quantified using the advanced method for accurate, robust, and efficient spectral fitting. FWM, frontal white matter; IGM, insular gray matter; OGM, occipital gray matter; PWM, parietal white matter; TGM, thalamic gray matter.
metabolites. With 7-T MRS, it has been shown that lowconcentration metabolites such as glycine can be detected (20), and contiguous metabolites such as Cr and phosphocreatine can be reliably differentiated from one another (5). The evaluation of the distribution of these low-concentration metabolites could be a focus for further applications. In addition, future plans include the optimization of the pointresolved spectroscopic sequence to obtain shorter repetition and echo times to enable high-efficiency measurement of J-modulating compounds (glutamate, glutamine) and short– echo time compounds such as macromolecules. CONCLUSIONS Regional differences between cerebral metabolite distributions were confirmed. These data can therefore provide a basis for further clinical applications of proton-based cerebral MRS in diseases that alter cerebral metabolites, including ischemia, tumors, and inflammatory or psychiatric disorders. REFERENCES 1. Inglese M, Rusinek H, George IC, et al. Global average gray and white matter N-acetylaspartate concentration in the human brain. Neuroimage 2008; 41:270–276. 2. Noworolski SM, Nelson SJ, Henry RG, et al. High spatial resolution 1H-MRSI and segmented MRI of cortical gray matter and subcortical white matter in three regions of the human brain. Magn Reson Med 1999; 41: 21–29. 3. Pouwels PJ, Frahm J. Regional metabolite concentrations in human brain as determined by quantitative localized proton MRS. Magn Reson Med 1998; 39:53–60. 4. Schuff N, Ezekiel F, Gamst AC, et al. Region and tissue differences of metabolites in normally aged brain using multislice 1H magnetic resonance spectroscopic imaging. Magn Reson Med 2001; 45:899–907. 5. Tkac I, Andersen P, Adriany G, et al. In vivo 1H NMR spectroscopy of the human brain at 7 T. Magn Reson Med 2001; 46:451–456.
6. Mazzoni LN, Belli G, Ginestroni A, et al. Computation of brain metabolite ratios in single-voxel proton MR spectroscopy: comparison between semiautomatic and automatic software. Radiol Med 2010; 115: 125–132. 7. Torriani M, Thomas BJ, Halpern EF, et al. Intramyocellular lipid quantification: repeatability with 1H MR spectroscopy. Radiology 2005; 236: 609–614. 8. Vanhamme L, van den Boogaart A, Van Huffel S. Improved method for accurate and efficient quantification of MRS data with use of prior knowledge. J Magn Reson 1997; 129:35–43. 9. Arnold DL, de Stefano N, Matthews PM, et al. N-acetylaspartate: usefulness as an indicator of viable neuronal tissue. Ann Neurol 2001; 50:823–825. 10. Soares DP, Law M. Magnetic resonance spectroscopy of the brain: review of metabolites and clinical applications. Clin Radiol 2009; 64:12–21. €fer LC, et al. 8-channel transmit/receive head coil 11. Orzada S, Kraff O, Scha for 7 T human imaging using intrinsically decoupled strip line elements with meanders. Presented at: ISMRM 17th Scientific Meeting & Exhibition; Honolulu, HI; 2009. 12. Frahm J, Bruhn H, Gyngell ML, et al. Localized proton NMR spectroscopy in different regions of the human brain in vivo. Relaxation times and concentrations of cerebral metabolites. Magn Reson Med 1989; 11:47–63. 13. Wang Y, Li SJ. Differentiation of metabolic concentrations between gray matter and white matter of human brain by in vivo 1H magnetic resonance spectroscopy. Magn Reson Med 1998; 39:28–33. 14. Wiedermann D, Schuff N, Matson GB, et al. Short echo time multislice proton magnetic resonance spectroscopic imaging in human brain: metabolite distributions and reliability. Magn Reson Imaging 2001; 19: 1073–1080. 15. Hetherington HP, Mason GF, Pan JW, et al. Evaluation of cerebral gray and white matter metabolite differences by spectroscopic imaging at 4.1T. Magn Reson Med 1994; 32:565–571. 16. Otazo R, Mueller B, Ugurbil K, et al. Signal-to-noise ratio and spectral linewidth improvements between 1.5 and 7 Tesla in proton echo-planar spectroscopic imaging. Magn Reson Med 2006; 56:1200–1210. 17. Barker PB, Hearshen DO, Boska MD. Single-voxel proton MRS of the human brain at 1.5T and 3.0T. Magn Reson Med 2001; 45:765–769. 18. Gonen O, Gruber S, Li BS, et al. Multivoxel 3D proton spectroscopy in the brain at 1.5 versus 3.0 T: signal-to-noise ratio and resolution comparison. AJNR Am J Neuroradiol 2001; 22:1727–1731. 19. Vaughan JT, Garwood M, Collins CM, et al. 7T vs. 4T: RF power, homogeneity, and signal-to-noise comparison in head images. Magn Reson Med 2001; 46:24–30. 20. Gambarota G, Mekle R, Xin L, et al. In vivo measurement of glycine with short echo-time 1H MRS in human brain at 7 T. Magma 2009; 22:1–4.
587