UHF MRSI: SNR, Speed & Resolution
Michal Považan1

1Department of Radiology and Radiological Science, School of Medicine, Johns Hopkins University, Baltimore, MD, United States


The temporal and spatial resolution of an MR experiment is always influenced by the available signal-to-noise ratio (SNR). In general, SNR increases with a higher static magnetic field (B0). Magnetic resonance spectroscopic imaging may benefit from the ultra-high field, however novel approaches are necessary to overcome the technical challenges that arise at such high magnetic field strengths. In this talk we focus on the specifics of UHF MRSI and present the most recent MRSI methods where the SNR gain can be traded off for higher spatial or temporal resolution.

Target audience

MR physicists, clinicians and technologists who are interested in learning about the specifics of MRSI at fields above 3T as well as state-of-the-art MRSI techniques


· Understanding of the major advantages and drawbacks of MRSI methods at ultra-high fields

· Get an overview of the most recent MRSI methods

· Explore a world beyond vendor-provided MRSI sequences


Magnetic resonance spectroscopic imaging (MRSI) is a non-invasive technique that merges the principles of magnetic resonance spectroscopy (MRS) and magnetic resonance imaging (MRI). It provides valuable insights into biochemical processes inside the human (or animal) body. The temporal (i.e. speed) and spatial resolution of an MR experiment is always influenced by the available signal-to-noise ratio (SNR). These three factors are closely linked together. Every MRSI experiment is a tradeoff between them in order to achieve short measurement times, high spatial resolution and sufficient SNR for quantification. In general, SNR increases with a higher static magnetic field (B0) 1. In addition, larger chemical shift dispersion and a reduction in J-coupling of strongly coupled systems is observed at UHF 2. Therefore, a recent trend towards installation of ultra-high field (UHF, i.e. B0 ≥7T) MR systems may be beneficial for MRSI. However, novel approaches are necessary to overcome the technical challenges that arise at such a high magnetic field strength 3.

Major challenges at UHF

UHF has introduced many challenges that need to be addressed. Specific absorption rate is proportional to B02, which considerably affects the maximum available pulse RF power. Hence, novel RF pulses and pulse sequences have to be developed for UHF applications to minimize the power deposition. T2 relaxation times are shorter at higher field strengths for the majority of metabolites. This penalizes sequences with longer echo time or acquisition delay (e.g. PRESS, LASER), as they suffer from significant signal loss at UHF. At UHF, B0, and B1+ (i.e., transmit B1) inhomogeneities are more pronounced. The increase in wavelength leads to a spatially variable B1 field due to standing wave- and skin effects. Hence, it is more difficult to achieve the desired flip angles over a large volume or in some brain regions. Susceptibility effects are also enhanced, which cause higher B0 inhomogeneities. B0 shimming procedures are therefore more demanding compared to lower fields. In addition, chemical shift displacement (CSD) is increased due to the higher chemical shift dispersion; in fact one of the greatest advantages of UHF. In extreme cases, the observed metabolites are detected from regions that do not overlap at all. This biases the spectral quantification.

UHF MRSI techniques

Due to substantial B1+ field inhomogeneities and huge CSD, the work-horse of spectroscopic localization, PRESS, is not feasible at UHF. The replacement of conventional refocusing pulses by a pair of adiabatic broadband refocusing pulses in semi-LASER represents a robust UHF approach 4,5. STEAM, despite 50% SNR reduction, may be utilized as well 6,7. Novel, numerically optimized pulses have been designed and implemented to overcome low bandwidth and to reduce CSD of PRESS 7,8. The biggest drawbacks of the above mentioned methods are the relatively long TE and the 3D rectangular pre-selection that excludes cortical areas of brain 9. Otazo et al. 10, in one of the first papers on MRSI at 7T, addressed both of these issues by a slice selective spin echo sequence with echo-planar read-out (EPSI). The search for robust UHF localization with low SNR, no pre-selection and a negligible CSD led to a slice-selective free induction decay (FID) MRSI 7,11,12. The lack of refocusing pulses allows ultra-short TEs which minimizes the SNR loss due to T2* decay and J-evolution. The gain in SNR can be further traded for higher spatial or temporal resolution.

B0 shimming for UHF MRSI

The increased B0 inhomogeneity limits the brain coverage of MRSI at UHF. Advanced methods such as dynamic B0 and RF shimming 13 and high-order and degree spherical harmonic shimming 14,15 may substantially improve UHF MRSI data quality.


Static magnetic field (B0) of 7T and higher has allowed the increase of spatial and temporal resolution of MRSI. However, it has also introduced new challenges that required a development of new MRSI methods tailored for UHF.


No acknowledgement found.


1. Trattnig, S. et al. Clinical applications at ultrahigh field (7 T). Where does it make the difference? NMR Biomed. 29, 1316–1334 (2016).

2. Uğurbil, K. et al. Ultrahigh field magnetic resonance imaging and spectroscopy. Magn. Reson. Imaging 21, 1263–1281 (2003).

3. Henning, A. Proton and multinuclear magnetic resonance spectroscopy in the human brain at ultra-high field strength: A review. Neuroimage 168, 181–198 (2017).

4. Scheenen, T., Heerschap, A. & Klomp, D. Towards 1 H-MRSI of the human brain at 7T with slice-selective adiabatic refocusing pulses. Magn. Reson. Mater. Physics, Biol. Med. 95–101 (2008). doi:10.1007/s10334-007-0094-y

5. Donadieu, M. et al. Metabolic Voxel-Based Analysis of the Complete Human Brain Using Fast 3D-MRSI : Proof of Concept in Multiple Sclerosis. (2016). doi:10.1002/jmri.25139

6. Chadzynski, G. L. et al. In vivo proton magnetic resonance spectroscopic imaging of the healthy human brain at 9.4 T: initial experience. Magn. Reson. Mater. Physics, Biol. Med. 28, 239–249 (2014).

7. Henning, A., Fuchs, A., Murdoch, J. B. & Boesiger, P. Slice-selective FID acquisition, localized by outer volume suppression (FIDLOVS) for (1)H-MRSI of the human brain at 7 T with minimal signal loss. NMR Biomed. 22, 683–96 (2009).

8. Balchandani, P., Pauly, J. & Spielman, D. Interleaved narrow-band PRESS sequence with adiabatic spatial-spectral refocusing pulses for 1H MRSI at 7T. Magn. Reson. Med. 59, 973–979 (2008).

9. Henning, A. et al. SELOVS: Brain MRSI localization based on highly selective T1- and B1-insensitive outer-volume suppression at 3T. Magn. Reson. Med. 59, 40–51 (2008).

10. Otazo, R., Mueller, B., Ugurbil, K., Wald, L. & Posse, S. Signal-to-noise ratio and spectral linewidth improvements between 1.5 and 7 tesla in proton echo-planar spectroscopic imaging. Magn. Reson. Med. 56, 1200–1210 (2006).

11. Boer, V. O., Klomp, D. W. J., Juchem, C., Luijten, P. R. & De Graaf, R. A. Multislice 1H MRSI of the human brain at 7 T using dynamic B0 and B1 shimming. Magn. Reson. Med. 68, 662–670 (2012).

12. Bogner, W., Gruber, S., Trattnig, S. & Chmelík, M. High resolution mapping of human brain metabolites by free induction decay 1H MRSI at 7 T. NMR in Biomed 873–882 (2012). doi:10.1002/nbm.1805

13. Boer, V. O., van de Bank, B. L., van Vliet, G., Luijten, P. R. & Klomp, D. W. J. Direct B0 field monitoring and real-time B0 field updating in the human breast at 7 Tesla. Magn. Reson. Med. 67, 586–91 (2012).

14. Chang, P., Nassirpour, S. & Henning, A. Modeling real shim fields for very high degree (and order) B0 shimming of the human brain at 9.4 T. Magn. Reson. Med. 79, 529–540 (2018).

15. Nassirpour, S., Chang, P., Fillmer, A. & Henning, A. A comparison of optimization algorithms for localized in vivo B0 shimming. Magn. Reson. Med. 0, 1–12 (2017).

Proc. Intl. Soc. Mag. Reson. Med. 26 (2018)