Optimization of static magnetic field homogeneity in the human and animal brain in vivo

https://doi.org/10.1016/j.pnmrs.2008.04.001Get rights and content

Introduction

Samples studied with conventional magnetic resonance (MR) techniques are subjected to a polarizing static magnetic field B(r)=B0zˆ. Ideally, this magnetic field is assumed to be spatially homogeneous over the sample volume. However, the presence of any object of finite magnetic susceptibility (i.e. a sample) will inevitably perturb this magnetic field and provide the Larmor frequencies of MR-sensitive spins with an unwanted spatial dependence. These susceptibility-induced B0 perturbations, which are a fundamental result of Maxwell’s electromagnetic field theory, scale roughly linearly in magnitude with applied B0 field strengths.

B0 inhomogeneity degrades the signal-to-noise ratio (SNR) of all MR measurements. While Hahn spin-echoes can be used to refocus B0 inhomogeneity at one instantaneous temporal point, the remainder of a spin-echo is tempered by B0 inhomogeneity-induced signal relaxation (commonly referred to as T2*). Furthermore, many techniques cannot utilize spin-echo procedures, and are thus susceptible to the full evolution of T2* relaxation.

The spatial variation of Larmor frequencies within a spectroscopic voxel will broaden and distort spectral lineshapes. Standard spectroscopic techniques such as frequency-selective resonance suppression and spectral editing are easily degraded by these effects.

MR images are susceptible to the same B0-induced SNR deterioration as spectroscopic acquisitions. Additionally, they can further suffer from spatial distortion in regions of high B0 inhomogeneity. In particular, commonly utilized rapid imaging strategies such as steady-state free precession (SSFP), spiral, and echo-planar imaging (EPI) are often compromised by B0 inhomogeneity.

Broadly speaking, B0 shimming refers to the optimized application of external magnetic fields to compensate unwanted inhomogeneity of the B0 magnetic field. For applications that require high-degrees of B0 homogeneity, fine-tuned shimming has historically been accomplished using sets of dedicated electromagnets. These electromagnet coils, or ‘active shims’, can be adjusted on a subject-specific basis.

This review is focused on high-field B0 shimming of the brain, specifically within small rodents and humans. B0 field perturbations within the brain are particularly prominent near the air-tissue interfaces at the sinus and auditory cavities. The recent increases of B0 field strengths used in both clinical and research MR systems necessitates maximal utility of conventional shim technology. For some applications, this conventional technology cannot adequately homogenize designated shim volumes (particularly larger volumes). Therefore, along with the development of automated optimization protocols for conventional active shim systems, recent investigations have explored alternative approaches to shim hardware design.

A theoretical treatment of B0 inhomogeneity and its effects on magnetic resonance acquisitions are presented in the next section. Section 3 then presents the techniques utilized in MR-based mapping of static magnetic fields, which is now a standard tool in automated shimming methods. The hardware and methods utilized in room-temperature shimming (also commonly referred to as ‘active’ shimming) are developed in Section 4. The challenges of optimizing B0 homogeneity over extended volumes are introduced in Section 5, followed by a discussion of the methods and capabilities of Dynamic Shim Updating (DSU) of room-temperature shims. Finally, recent novel approaches to shimming that deviate from any previous technological methodologies, such as local active shimming and subject-specific passive shimming, are presented.

Section snippets

Theory

An in vivo MR system will typically have RMS B0 homogeneity of up to only about 50 parts-per-million (ppm) of the static magnetic field strength (≈6 kHz at 3T) over manufacturer-defined diameter spherical volumes (DSVs). Typical DSV diameters are roughly 45 cm for human systems and 5–6 cm for small animal systems. This inherent magnetic field inhomogeneity arises from a number of factors. First, it is extremely difficult to generate homogeneous current density distributions in superconducting

Principles

The basic principle of MRI-based B0 field mapping relies upon quadrature (complex) detection of the MR signal. Quadrature detection provides information on the angular phase (ϕ) of measured transverse magnetization. The phase of the measured magnetization advances with echo-time TE, according toϕ=ϕ0+ωTE,where ϕ0 is the initial phase of the magnetization imparted by RF excitation and ω is related to the local magnetic field offset by,ω=γB0+ωCS+γΔB0.Here, B0 is the applied static magnetic field, ω

Room temperature (RT) shimming

The zˆ-component of the static magnetic field in an MR experiment, BzT(r), is given byBzT(r)=B0+δB0(r)+BRTS(r)+BPS(r)+BS(r),where the superimposed contributions represent the ideal homogeneous superconducting-electromagnet field (B0), imperfections in this field (δB0), the room-temperature (RT) shim field (BRTS), the passive shim field (BPS), and the sample-induced field (BS). The object of shimming is to generate a pure B0 field, and therefore satisfy the conditionBS(r)+δB0(r)=-BRTS(r)+BPS(r).

Shimming over extended volumes

The methods presented in the previous section can sufficiently homogenize the B0 distribution over small volumes and are often adequate for single-voxel spectroscopy applications. However, for many applications the aforementioned methods and hardware cannot adequately homogenize extended volumes across the whole brain.

Fig. 8 shows the effects of increasing orders of shim inclusion in the whole-brain least-squares optimization of B0 field homogeneity in the human brain at 3T. It is clear from

Conclusion

B0 shimming remains an active field of technological development. Sections 2 Theory, 3 Magnetic field mapping, 4 Room temperature (RT) shimming have highlighted some of the primary considerations involved in such development. Utilizing the founding principles of magnetostatic physics, shim coil design, magnetic field mapping, and automated shim optimization, Section 5 presented the remaining challenges posed by B0 shimming of the brain. Advanced shim optimization methods addressing these

Acknowledgments

The authors would like to thank Dr. Jason Hsu, Dr. Gary Glover, Dr. Peter Jezzard, Dr. Peter van Gelderen, Dr. Michael Poole, and Dr. Richard Bowtell for contributing their results to this review. Further gratitude goes to Terence Nixon, Scott McIntyre, Peter Brown and Dr. Laura Sacolick for their assistance in developing and implementing aspects of the technology presented in this review. Previously unpublished results contained in this review were funded by NIH Grants R21-CA118503 and

Glossary

Active shimming
B0 shimming whereby sets of electromagnetic coils are used to shim in a subject-specific fashion
Diamagnetic materials
Materials that induce magnetic fields which oppose applied fields (negative magnetic susceptibility)
DSU: dynamic shim updating
Technique whereby RT shims are updated during an MR acquisition. Shims can be dynamically updated for specific spatial regions (slices or voxels) or in time (respiration, movement)
EPI: echo-planar imaging
Magnetic resonance imaging technique

First page preview

First page preview
Click to open first page preview

References (86)

  • Y. Zhao et al.

    J. Magn. Reson.

    (2005)
  • K. Ugurbil et al.

    Magn. Reson. Imaging

    (2003)
  • R. Cusack et al.

    Neuroimage

    (2002)
  • C.M. Collins et al.

    Magn. Reson. Imaging

    (2002)
  • D. Hoult et al.

    J. Magn. Reson. A

    (1994)
  • R. Gruetter et al.

    J. Magn. Reson.

    (1992)
  • J. van Vaals et al.

    J. Magn. Reson.

    (1990)
  • M. Garwood et al.

    J. Magn. Reson.

    (2001)
  • J.L. Wilson et al.

    Neuroimage

    (2003)
  • C. Juchem et al.

    J. Magn. Reson.

    (2006)
  • K.M. Koch et al.

    J. Magn. Reson.

    (2006)
  • C.N. Chen et al.

    Biomedical Magnetic Resonance Technology

    (2005)
  • G. Neuberth, US Patent (2005)...
  • J.D. Jackson

    Classical Electrodynamics

    (1999)
  • R. Salomir et al.

    Concepts Magn. Reson. B

    (2003)
  • J.P. Marques et al.

    Concepts Magn. Reson. B

    (2005)
  • K.M. Koch et al.

    Phys. Med. Biol.

    (2006)
  • D.A. Yablonskiy

    Magn. Reson. Med.

    (1998)
  • S. Ogawa et al.

    Magn. Reson. Med.

    (1990)
  • K.K. Kwong et al.

    Proc. Natl. Acad. Sci. USA

    (1992)
  • P.A. Bandettini et al.

    Magn. Reson. Med.

    (1992)
  • Y. Zur et al.

    Magn. Reson. Med.

    (1988)
  • N.K. Bangerter et al.

    Magn. Reson. Med.

    (2004)
  • E. Yudilevich et al.

    IEEE Trans. Med. Imaging

    (1987)
  • P. Jezzard et al.

    Magn. Reson. Med.

    (1995)
  • J.L. Andersson, M. Richter, W. Richter, S. Skare, R.G. Nunes, M. Robson, T. Behrens, in: Proceedings of the 15th...
  • N. Bloembergen et al.

    Phys. Rev.

    (1948)
  • R.A. de Graaf et al.

    Magn. Reson. Med.

    (2006)
  • S. Michaeli et al.

    Magn. Reson. Med.

    (2002)
  • E.L. Hahn

    Phys. Rev.

    (1950)
  • D.A. Yablonskiy et al.

    Magn. Reson. Med.

    (1994)
  • R.A. de Graaf

    In Vivo NMR Spectroscopy: Principles and Techniques

    (1998)
  • A. Haase et al.

    Phys. Med. Biol.

    (1985)
  • M. Hedley et al.

    Magn. Reson. Med.

    (1992)
  • S.M. Song et al.

    IEEE Trans. Image Process.

    (1995)
  • S. Chavez et al.

    IEEE Trans. Med. Imaging

    (2002)
  • M. Jenkinson

    Magn. Reson. Med.

    (2003)
  • R.J. Reber et al.

    Magn. Reson. Med.

    (1998)
  • D. Kim et al.

    Magn. Reson. Med.

    (2002)
  • L.M. Klassen et al.

    Magn. Reson. Med.

    (2004)
  • J.F. Schenck

    Med. Phys.

    (1996)
  • J.A. Hopkins et al.

    Magn. Reson. Med.

    (1997)
  • BioImage, Bioimaging Sciences, Yale University, New Haven, CT,...
  • Cited by (57)

    • Bench to bore ramifications of inter-subject head differences on RF shimming and specific absorption rates at 7T

      2022, Magnetic Resonance Imaging
      Citation Excerpt :

      Simulation work focusing on uniform models and optimum distance for efficiency has already been done [17] but the human head is far from uniform and population heterogeneity may lead to very different scattering parameters. Arrays should consider as many human models as possible as has been done with gradient coils [44,45]. Hybridized approaches for coil design including transmit homogeneity or SAR variation across many models could make RF coils more universal.

    • Boosting resolution in NMR spectroscopy by chemical shift upscaling

      2020, Analytica Chimica Acta
      Citation Excerpt :

      A number of experimental methods and data processing approaches have been developed to enhanced spectral resolution in NMR spectroscopy. Advanced shimming techniques were designed for better field homogeneity [9,10]. There are also many methods aimed at recovering high-resolution information from inhomogeneous fields [11–13].

    View all citing articles on Scopus
    1

    Tel.: +414 218 7991; fax: +262 521 6589.

    View full text