Single-Voxel Spectroscopy at 7T & Beyond: From Animal to Human
Dinesh Deelchand1
1University of Minnesota, United States

Synopsis

This lecture focuses on the pros and cons of utilizing single-voxel proton MR spectroscopy at ultra-high fields (UHF) of 7T and beyond in both human and animal brains. Advantages include higher signal-to-noise ratio, higher spectral dispersion i.e. less overlap between metabolites and these benefits lead to improved quantification of metabolites. However in human brain, going to UHF is associated with B1 inhomogeneity and increased RF power requirements and these can be mitigated by using dielectric pads or B1 shimming techniques. In addition, relaxation times of metabolites and water tissue signals change as B0 field increases.

Introduction

Proton magnetic resonance spectroscopy (1H MRS) is a non-invasive technique that enables measurement of the concentration of at least 18 metabolites in the human and animal brains [1-4]. MRS data can be acquired using single-voxel spectroscopy (SVS) or with magnetic resonance spectroscopic imaging (MRSI). SVS is the most commonly used technique due to high spectral quality (e.g. good localization, optimized B0 shim and water suppression). However, the method is limited to only single volume-of-interest (VOI). Commonly utilized SVS sequences are STEAM [5], PRESS [6], SPECIAL [4], LASER [7] and semi-LASER [8-9]. The benefits and issues associated with performing SVS acquisitions at ultra-high fields (UHF) of 7T and higher are discussed below.

Benefits of higher magnetic fields

  • Increased signal-to-noise ratio (SNR) especially in the time domain. However in the frequency domain, the SNR of NAA levels off with modest increase above 3 T in the human brain [10]. In rodent brain, the SNR of NAA levels off beyond 16 T [11].
  • Increased in spectral dispersion [12]. Less overlap between metabolites. For instance, glutamate and glutamine can be resolved beyond 7 T in both human and animal brains [13].
  • Simplification of J-coupled spectral patterns. Based on increased spectral linewidth (mostly from microscopic susceptibility effects) and dispersion with higher B0 field [2]. ·
  • Improved quantification precision of metabolite. This is reflected in the Cramer-Rao Lower bounds (CRLB). Several studies have reported decreased in CRLB from 3 or 4 T to 7 T [14-15]. For singlets, CRLB improves as ~sqrt(B0) [2]. For J-coupled metabolites (e.g. glutamate, glutamine, myo-inositol), the gain in quantification precision is much more and ranges from sqrt(B0) to B0 [2].

Issues associated with UHF

  • Shorter transverse (T2) relaxation times of water and metabolites. T2 decreases exponentially in the human brain [2].
  • Longer longitudinal (T1) relaxation times of water and metabolites. Modest increase in T1 values in human and animal brains [2].
  • RF power requirement: Minimize chemical shift displacement error (CSDE): need high-bandwidth RF pulses to cover the larger spectral dispersion. For refocusing pulses, gradient-modulated RF pulses such as FOCI [16] or GOIA [17-18] pulses can be utilized to reduce required B1 field.
  • RF field uniformity goes down. B1 issues especially in humans brain [19]. This can be mitigated by using B1 shimming routine when using multi-channel coils for transmit or dielectric pad.

Localization MRS sequences for UHF

  • Requirements: large RF bandwidth, small CSDE, short or relatively short echo-time (TE) sequences (due to shorter T2 relaxation times).
  • STEAM or semi-adiabatic SPECIAL: these can achieve relatively short TE although SNR is halved compared with STEAM compared to spin-echo sequences.
  • LASER or semi-LASER: T2 relaxation times of water and metabolites are lengthened due to pairs of adiabatic RF pulses in these sequences [20]. In addition, J-modulation is suppressed [20]. This in turn means that relatively long TE MR spectrum resembles short-TE spectra acquired with STEAM.

Acknowledgements

No acknowledgement found.

References

[1] Pfeuffer et al. J Magn Reson 1999

[2] Deelchand et al. J Magn Reson 2010

[3] Hong et al. Magn Reson Med 2011

[4] Mlynárik et al. J Magn Reson 2008

[5] Frahm et al. J Magn Reson 1969

[6] Bottomley Ann NY Acad Sci 1987

[7] Garwood & DelaBarre J Magn Reson 2001

[8] Scheenen et al. Magn Reson Med 2008

[9] Öz & Tkáč Magn Reson Med 2011

[10] Deelchand et al. Magn Reson Med 2014

[11] de Graaf et al. Magn Reson Med 2006

[12] Gruetter et al. J Magn Reson 1998

[13] Tkac et al. Magn Reson Med 2001

[14] Mekle et al. Magn Reson Med 2009

[15] Tkac et al. Magn Reson Med 2009

[16] Ordidge et al. Magn Reson Med 1996

[17] Tannus & Garwood NMR Biomed. 1997

[18] Andronesi et al. J Magn Reson. 2010

[19] van de Moortele et al. Magn. Reson. Med. 2005

[20] Deelchand et al. Magn. Reson. Med. 2015

Proc. Intl. Soc. Mag. Reson. Med. 28 (2020)