3015

Qualitative Comparison between In Vivo J-Resolved Semi-LASER at 3 T and 9.4 T

Saipavitra Venkateshwaran Murali Manohar¹, Ioannis Angelos Giapitzakis¹, Tamas Borbath¹, Matti Gaertner², and Anke Henning^1,3

¹High Field Magnetic Resonance, Max Planck Institute for Biological Cybernetics, Tuebingen, Germany, ²Department of Psychiatry, Charité - Universitätsmedizin Berlin, ³Institute of Physics, Ernst-Moritz-Arndt University Greifswald, Greifswald, Germany

Synopsis

J-resolved semi-LASER with maximum-echo sampling is optimized at 9.4T and compared with the same implementation at 3T in terms of SNR and spectral resolution. SODA scheme is appreciated for the sequence rather than the MC scheme. SNR at 9.4T (t₁ steps: 85) was approximately 5.8 times greater than at 3T (t₁ steps: 100) and strongly coupled peaks are well-resolved. However, the trade-off between SNR and spectral resolution is explained as lactate (1.32 ppm), a weakly-coupled metabolite, is better resolved at 3T. Higher band-width AFP pulses helped in almost vanishing the J-refocused peaks which made the J-resolved peaks clearly distinguishable. A few interesting downfield peaks and the doublet of NAA (7.82ppm) are observed.

Purpose

Enhancing the signal-to-noise ratio(SNR) and the spectral resolution have always been the ambition in the field of in vivo ¹H magnetic resonance spectroscopy. Spectroscopic studies at ultra-high field benefit from both increased spectral resolution and improved SNR. A distinct method to reduce spectral signal overlap is 2D spectroscopy, which spreads the spectral information along two frequency axes^1-5. Exploiting these two complimentary approaches, a J-resolved semi-LASER sequence with maximum-echo sampling(MES) scheme was optimized for the application in human brain at 9.4T. Also a qualitative comparison between J-resolved semi-LASER at 3T¹ and 9.4T in terms of SNR and spectral resolution was done.

Methods

J-resolved semi-LASER experiments were performed on a Siemens Magnetom 9.4T and a Siemens Trio 3T whole-body MRI scanner(Erlangen, Germany). A voxel of 2x2x2 cm³ was chosen in the occipital lobe with a mixed gray and white matter content at 9.4T and in the pre-frontal cortex at 3T. Four-channel transceiver coil⁶ was used at 9.4T and was shimmed using the FASTEST MAP⁷. Power calibration was performed for the selected voxel. At 3T and 9.4T, the shortest TE was 24ms, the step size was ∆TE=2ms. 50 or 85 echo time increments were recorded at 9.4T whereas at 3T 100 steps^5,8 were recorded. Bandwidth in the direct dimension was 4000Hz and 2000Hz at 9.4T and 3T, respectively. The sampling points were 4096 in the direct dimension. MES scheme was implemented(Figure 1). Metabolite cycling(MC)⁹ scheme was tested for the sequence with 8 and 16 phase-cycling steps and the resulting spectral quality was compared to experiments with SODA¹⁰ water suppression(Short duration water suppression using optimized ﬂip angles). Also a non-water suppressed experiment was performed with otherwise identical scan parameters in order to allow for Eddy current correction¹¹.The sequence was optimized by performing scans on phantom and 8 healthy volunteers and later 7 healthy volunteers participated in the study at 3T(n=1), 9.4T with 50(n=5) and 85 steps(n=1), respectively, all of them gave informed signed consent as approved by the local ethics board.

Results

The MES scheme gave the peak tails a tilt in the 2D spectra and therefore avoided overlap of spectral peaks and peak tails from high intensity metabolites and water. SODA water suppression, yielded a suppression factor of 99.47% at 9.4T and 99.23% at 3T, gave minimal ghosting artifacts whereas MC gave stronger ghosting artifacts in the 2D spectra for 8 phase-cycling steps, which reduced with 16-step phase cycle. Due to the high pulse bandwidth and minimal chemical shift displacement artifact, J-refocused peaks were barely visible in either the 2D spectra at 3T or at 9.4T(Figure 3,4). J-resolved peaks could be well-distinguished at 9.4T(Figure 2,3) while 3T J-resolved semi-LASER data exhibited a stronger spectral overlap. However, the lactate peak at 1.32 ppm could clearly be resolved at 3T but not at 9.4T due to a lower number of sampling points in the indirect dimension(Figure 3,4) which represented a trade-off between SNR and spectral resolution at 9.4T. At only 9.4T it was possible to observe some of the down-field peaks¹² like NAA(7.82ppm) and hCs(8.08ppm). The doublet of NAA at 7.82 ppm was prominently noticeable at 9.4T(Figure 2).

Discussion

J-resolved semi-LASER with MES scheme was investigated for the first time at 9.4T and compared with 3T results. Compared to previous 1D MRS studies at 9.4T¹³ low number of averages per TE in J-resolved MRS restricts the number of possible phase cycling steps(herein 8), which appeared to be insufficient to suppress ghosting artifacts introduced by metabolite cycling. Nevertheless, SODA¹⁰ provided an excellent water suppression despite B₁⁺ inhomogeneity at high field strengths and at 3T. Unwanted coherence pathways were suppressed during the SODA water suppression scheme using the most suitable combination of gradients. The compromise between SNR and spectral resolution of the J-resolved spectra was investigated. The weakly coupled metabolites like lactate could be better resolved at 3T rather than at 9.4T because of a larger TE range due to the Carr-Purcell behavior(CP)^14,15 of the semi-LASER sequence(Figure 4). Though 9.4T showed a higher SNR(Table 1) and strongly-coupled peaks such as NAA_asp and mI were well-resolved, T₂ relaxation times are shorter at 9.4T¹⁶ than 3T. So, optimizing the TE range to avoid loss of signal while providing sufficient time for J-evolution is required in order to achieve both higher SNR and well-resolved weakly-coupled peaks. The appearance of the downfield peaks showed that J-resolved semiLASER at 9.4T could be a potential method to help assigning unknown peaks by understanding their J-coupling properties, although the optimal TE range for downfield MRS should be smaller as their T₂ values are shorter¹⁷.

Acknowledgements

This work was supported by the Horizon 2020 CDS-QUAMRI grant. The authors thank Simone Grimm, Evgeniya Kirilina, Yan Fan, Anna Stippl, Ana Herrera from Charité - Universitätsmedizin Berlin for help in this work.

References

1. Lin M, Kumar A, Yang S. Two-Dimensional J-Resolved LASER and Semi-LASER Spectroscopy of Human Brain. Magn Reson Med 2014; 71(3):911-920.

2. Thomas M, Yue K, Binesh N, et al. Localized two-dimensional shift correlated MR spectroscopy of human brain. Magn Reson Med 2001; 46:58–67.

3. Andronesi O, Ramadan S, Mountford C, Sorensen A. Low-power adiabatic sequences for in vivo localized two-dimensional chemical shift correlated MR spectroscopy. Magn Reson Med 2010; 64:1542–1556.

4. Lange T, Trabesinger A, Schulte R, Dydak U, Boesiger P. Prostate spectroscopy at 3 Tesla using two-dimensional S-PRESS. Magn Reson Med 2006; 56:1220–1228.

5. Schulte R F, Lange T, Beck J, Meier D, Boesiger. Improved two-dimensional J-resolved spectroscopy. NMR Biomed 2006; 19(2): 264-270.

6. Pfrommer A, Avdievich N and Henning A (May-12-2014): Four Channel Transceiver Array for Functional Magnetic Resonance Spectroscopy in the Human Visual Cortex at 9.4 T, Joint Annual Meeting ISMRM-ESMRMB 2014, Milano, Italy.

7. Gruetter R, Tkac I. Field mapping without reference scan using asymmetric echo-planar techniques. Magn Reson Med 2000; 43(2):319-23.

8. Fuchs A, Boesiger P, Schulte RF, Henning A. ProFit revisited. Magn Reson Med. 2014; 71(2):458-68.

9. Giapitzakis IA, Nassirpour S, Avdievich N, Kreis R, Henning A. Metabolite cycled single voxel 1H spectroscopy at 9.4 T. 23rd Annual Meeting and Exhibition of ISMRM 2015.

10. Giapitzakis IA, Nassirpour S, Henning A. Short duration water suppresion using optimised flip angles (SODA) at ultra high fields.32nd Annual Scientific Meeting ESMRMB 2015.

11. Klose U. In vivo proton spectroscopy in presence of eddy currents. Magn Reson Med. 1990; 14(1):26-30.

12. Govindaraju V, Young K, Maudsley AA. Proton NMR chemical shifts and coupling constants for brain metabolites. NMR Biomed. 2000; 13(3):129-53.

13. Giapitzakis IA, Shao T, Avdievich N, Kreis R, Henning A. Metabolite cycled semi-LASER in human brain at 9.4T: In-vivo results. ISMRM Workshop on MR Spectroscopy, Lake Constance 2016.

14. Allerhand A, Thiele E. Analysis of Carr—Purcell Spin-Echo NMR Experiments on Multiple-Spin Systems. II. The Effect of Chemical Exchange. J Chem Phys. 1966; 45(3):902-16.

15. Deelchand DK, Henry PG, Marjanska M. Effect of carr-purcell refocusing pulse trains on transverse relaxation times of metabolites in rat brain at 9.4 Tesla. Magn Reson Med. 2015; 73(1):13-20.

16. Deelchand DK, Van de Moortele PF, Adriany G, Iltis I, Andersen P, Strupp JP, Vaughan JT, Ugurbil K, Henry PG. In vivo 1 H NMR spectroscopy of the human brain at 9.4 T: initial results. J Magn Reson. 2010; 206(1):74-80.

17. Fichtner ND, Henning A, Zoelch N, Boesch C, Kreis R. Eludication of the Downfield Spectrum of Human Brain at 7 T Using Multiple Inversion Recovery Delays and Echo Times. Magn Reson Med 2016.

18. Scheenen TW, Klomp DW, Wijnen JP, Heerschap A. Short echo time 1H-MRSI of the human brain at 3T with minimal chemical shift displacement errors using adiabatic refocusing pulses. Magn Reson Med. 2008; 59(1):1-6.

19. Edden RAE, Barker P B. If J doesn’t evolve, it won’t J-resolve: J-PRESS with bandwidth-limited refocusing pulses. Magn Reson Med 2011; 65(6): 1509-14.

Figures

Figure 1: Representative pulse sequence of SODA water suppression followed by J-resolved semi-LASER where the acquisition started immediately after the final gradient (MES). It consists of a slice-selective excitation 90^o RF pulse followed by two pairs of AFP pulses. The time interval between the last two AFP pulses was increased by ∆TE = 2 ms creating the indirect dimension. Semi-LASER¹⁸ was chosen for the J-resolved experiment as it has smaller chemical shift displacement effect (CSDE) (Cr-water=15%, Cr-fat=0.9%, since Cr was set as reference) due to a remarkable localization wherein it uses high bandwidth AFP pulses.

Figure 2: a) Three dimensional view of a logarithmic real plot of the spectra at 9.4T recorded with 50 TE increments in the t₁ dimension showing both the upfield and downfield peaks. b) Zoomed three dimensional view (top) of the downfield range where the NAA (N-acetyl aspartate) splitting is prominent at 7.82 ppm, and related 2D view (bottom) of the downfield spectra showing the NAA doublet, hCs (Homocarnosine) and other yet unassigned peaks.

Figure 3: a) 2D spectra (Logarithmic real plot) acquired at 9.4 T with 85 steps in the indirect dimension. b) Zoomed view of well-resolved Asp, NAA_asp and NAAG_asp peaks; J-refocused peaks¹⁹ are largely absent because of the high bandwidth (approximately 8000Hz) excitation and AFP refocusing pulses used at 9.4T as higher bandwidth reduces the signal from the J-refocused peaks by reducing the chemical shift displacement artifact.

Figure 4: 2D spectra (logarithmic real plot) acquired at 3T where the lactate peak at 1.32 ppm is better resolved than at 9.4T. Due to the CP behavior of the J-resolved semi-LASER sequence, there is inhibition to J-evolution when $$$\tau. \sqrt{\delta^2+J^2}<<1$$$ where $$$\tau$$$ is the time interval between the two AFP pulses, $$$\delta$$$ and J are the chemical shift displacement and J-coupling constant respectively. J-evolution becomes negligible when the last AFP pulse is a part of the CP train. Therefore, for TE_last=224ms at 3T, lactate J-resolved well, while TE_last=194ms at 9.4T (85 steps) it did not J-evolve fully.

Table 1: SNR (with respect to NAA peak at 2.008 ppm) comparison for data from 3T (with 100 sampling points in the t₁dimension) and 9.4T (MC: 8 phase-cycling steps, SODA: 50 and 85 sampling points in t₁ dimension). SNR was measured before any filtering in both dimensions. MC has an apparent poor SNR compared to SODA data at 9.4T due to the pronounced ghosting artifact that biases SNR estimation.

Proc. Intl. Soc. Mag. Reson. Med. 25 (2017)

3015