Elsevier

NeuroImage

Volume 182, 15 November 2018, Pages 8-38
NeuroImage

Microstructural imaging of the human brain with a ‘super-scanner’: 10 key advantages of ultra-strong gradients for diffusion MRI

https://doi.org/10.1016/j.neuroimage.2018.05.047Get rights and content

Highlights

  • Improved estimation of axon diameter in vivo.

  • Better fitting of complex microstructure models.

  • Improved characterisation of tumours.

  • Facilitation of multi-contrast MRI experiments.

Abstract

The key component of a microstructural diffusion MRI ‘super-scanner’ is a dedicated high-strength gradient system that enables stronger diffusion weightings per unit time compared to conventional gradient designs. This can, in turn, drastically shorten the time needed for diffusion encoding, increase the signal-to-noise ratio, and facilitate measurements at shorter diffusion times. This review, written from the perspective of the UK National Facility for In Vivo MR Imaging of Human Tissue Microstructure, an initiative to establish a shared 300 mT/m-gradient facility amongst the microstructural imaging community, describes ten advantages of ultra-strong gradients for microstructural imaging. Specifically, we will discuss how the increase of the accessible measurement space compared to a lower-gradient systems (in terms of Δ, b-value, and TE) can accelerate developments in the areas of 1) axon diameter distribution mapping; 2) microstructural parameter estimation; 3) mapping micro-vs macroscopic anisotropy features with gradient waveforms beyond a single pair of pulsed-gradients; 4) multi-contrast experiments, e.g. diffusion-relaxometry; 5) tractography and high-resolution imaging in vivo and 6) post mortem; 7) diffusion-weighted spectroscopy of metabolites other than water; 8) tumour characterisation; 9) functional diffusion MRI; and 10) quality enhancement of images acquired on lower-gradient systems. We finally discuss practical barriers in the use of ultra-strong gradients, and provide an outlook on the next generation of ‘super-scanners’.

Introduction

This article reviews the benefits of incorporating an ultra-strong gradient system (Setsompop et al., 2013), optimized for diffusion magnetic resonance imaging (dMRI), into an MRI scanner for imaging the human brain. It is written from the perspective of the National Facility for In Vivo MR Imaging of Human Tissue Microstructure, an initiative in the United Kingdom to establish an MRI system with ultra-strong (300 mT/m) gradients as a shared facility amongst the microstructural imaging community. We have selected ten key areas where we believe ultra-strong gradients can advance the field of microstructural imaging, beginning each section with motivation for advancing the field in each area, reviewing what has been achieved to date with more commonly-available gradient amplitudes, and then discussing the potential benefits and opportunities afforded by having access to ultra-strong gradients. In some cases, this is in providing marked improvements to measurements that are already attainable at lower gradient amplitude, and in others, it facilitates new measurements that are simply impractical at lower gradient amplitude. We then discuss practical challenges in making use of such a system, including issues of safety and engineering aspects, and attempt to highlight which limits are fundamental, and which just require engineering.

The main advantage is, of course, in providing a higher q-value/shorter echo time (TE) for a given b-value, and a higher signal-to-noise ratio (SNR) per unit b-value (see Fig. 1a). Shorter diffusion time acquisitions also become more practical as higher b-values can be achieved, and a wider range of b-values can be maintained across all diffusion times (Δ) (Fig. 1b).

Making measurements on such kit allows us to develop a truly translational pipeline, i.e., enabling measurements in humans that we could only previously make, in vivo, in animals on preclinical systems. Here, we firstly explore how stronger gradients provide improvements in characterising axon diameter, and microstructural parameters in general. Next we discuss the benefits for getting higher resolution characterisation of brain structure both in vivo and post mortem. We then discuss how, by looking at metabolites other than water, diffusion-weighted spectroscopy can provide deeper insights into tissue microstructure, and how ultra-strong gradients can accelerate developments in this area. While microstructural imaging offers huge potential benefits in a range of diseases, we select as an exemplar application the characterisation of tumours, and explore how stronger gradients may improve characterisation of tumour cellularity and other microstructural features. In addition to characterisation of structure, dMRI has been used to characterise brain function, with the study of rapid temporal changes in the microstructural milieu being an interesting application. Given the disparity of results in the literature on this topic, here we consider how ultra-strong gradients may help to cast further light on the relationship between neural activity and diffusion.

Despite the many potential advantages of ultra-strong gradients, we recognize that access to ultra-strong gradient systems is not likely to be wide-spread in the immediate future, and so we consider how having a small number of such systems, together with machine learning approaches, might still facilitate deeper insight into tissue microstructure beyond the immediate environs of the scanners themselves.

We then discuss practical barriers to making full use of such as system, considering the interplay of hardware and physiological safety limits on deployment of such gradient magnitudes, and consider how some of these issues may be addressed in future generations of ‘super-scanners’.

Section snippets

The challenge and its importance

One of the key features of the diffusion-weighted MR signal is its sensitivity to the size of pores with walls that are either fully restricting, or semi-permeable. The longer the characteristic length scale of the pores, the greater the mean displacement, and the lower the diffusion-weighted signal will be. In the brain, this capability underpins the tantalising prospect of estimating and mapping the distribution of axon diameters in each image voxel, since at least in white matter, the axonal

The challenge and its importance

To describe the complex diffusion process in tissue, a variety of approaches have been proposed that relate the signal to diffusion features or tissue microstructure, of which some have been described in the previous section (see also Panagiotaki et al., 2012; Assemlal et al., 2011; Tournier et al., 2011; and Novikov et al., 2018a, 2018b for further discussion). Different configurations of multi-compartment models have been evaluated in preclinical acquisitions (Panagiotaki et al., 2012) and in

The challenge and its importance

The majority of approaches developed to extract diffusion- or microstructural features from dMRI data, are built on the Stejskal-Tanner sequence (Stejskal and Tanner, 1965) (or ‘single diffusion encoding’, SDE). Diffusion tensor MRI (Basser et al., 1994), still being the most commonly applied SDE-based method to showcase sensitivity of dMRI to pathology, merely provides information on the average diffusion process within a voxel. Macroscopic DTI features such as fractional anisotropy (FA) and

The challenge and its importance

A variety of MRI contrasts have been used to characterise tissue beyond dMRI, such as relaxometry, susceptibility, and magnetization transfer. Multi-compartment models have become particularly popular in the attempt to provide a suitable description for each of these contrasts. Most commonly, such multi-compartment models are fitted to the signal from each MRI modality separately, and the joint and complementary information between contrasts therefore remains unexploited. In addition, important

The challenge and its importance

Until now, diffusion imaging methods, applied to the study of the living human brain, have been significantly limited by their own low spatial resolution. However, there are several important applications that could benefit from high resolution acquisitions. E.g. it has been shown that important cortical changes at the macro-scale level can occur in numerous psychiatric, neurodevelopmental and neurodegenerative disorders, (Goldman et al., 2009; Zielinski et al., 2014, Andrews et al., 2017;

The challenge and its importance

Post mortem MRI, and in particular diffusion imaging of whole brain specimens, has attracted interest in recent years for several reasons. Diffusion MRI has tremendous potential to contribute to our knowledge of neuroanatomy, including macroscopic connectivity, local fibre architecture, and patterns of microstructural features across the brain. However, in vivo anatomical investigations with dMRI are limited by achievable spatial resolution and diffusion contrast, with improvements in both

The challenge and its importance

While dMRI has predominantly focused on the diffusion of water molecules in tissue, the problem is that water is ubiquitous in tissue and exists in all compartments (e.g, intra-cellular and extra-cellular). It is this fundamental fact that makes it difficult to disentangle different tissue subcompartments (see the Sections on 'Measuring axon diameter' and 'Parameter estimation'). Diffusion-weighted Magnetic Resonance Spectroscopy (DW-MRS) uniquely enables specific characterisation of tissues

The challenge and its importance

The dMRI signal is potentially sensitive to a number of features of the tumour microenvironment that are relevant to diagnosis and assessment of treatment response. The promise of specific measurements of such properties is that they provide direct insight into key aspects of pathology that are only indirectly and, therefore potentially ambiguously, reflected using heuristic parameterisations such as ADC or kurtosis (e.g., Bourne, 2015; O'Connor et al., 2013). Getting access to such information

The challenge and its importance

Measuring neuronal activity non-invasively is fundamental to understanding the brain and its dysfunction occurring in disease. Direct measurements of the effects of neuronal current can be achieved with neurophysiological techniques, such as electro-encephalography (EEG) and magneto-encephalography (MEG). These methods, however, suffer from limited spatial resolution, and their sensitivity is typically restricted to cortical activity. Functional MRI (fMRI) based on blood-oxygenation

The challenge and its importance

In this article we have identified clearly the gains to be made by investing in ultra-strong diffusion gradient technology. However, as devices with this technology are not mass-produced, they incur considerable cost and are not currently widely available. A reasonable question, when reading an article such as this would be: “That's all very well, if you have access to such high performance equipment. However, most clinical/research imaging facilities do not. How then, can such technology be of

Practical issues for the use of ultra-strong gradients

The advantages of ultra-strong gradients for diffusion-based microstructure imaging have been clearly outlined in the previous sections. Here, we discuss the practical issues that set limits on the exploitation of ultra-strong gradients for different types of microstructure imaging studies. The Connectom scanner's ultra-strong gradient hardware can reach a maximum gradient amplitude of 300 mT/m (McNab et al., 2013). The maximum achievable slew-rate is 200 mT/m/ms, which is comparable with the

Conclusion

Even though diffusion encoding with strong gradients raises challenges, it is without question that with ongoing developments to better understand such issues, a microstructural ‘super-scanner’ provides considerable improvements to ‘conventional’ diffusion measurements and facilitates new measurements that are impractical at lower gradient amplitudes. New insights arising from this extended measurement space, together with practical experience on the challenges related to strong gradients, will

Acknowledgements

The UK National Facility for In Vivo MR Imaging of Human Tissue Microstructure was generously funded by the Engineering and Physical Sciences Research Council (EPSRC) (grant EP/M029778/1), and The Wolfson Foundation, to whom the authors give their thanks. DKJ is supported by a Wellcome Trust Investigator Award (096646/Z/11/Z) and a Wellcome Trust Strategic Award (104943/Z/14/Z). CMWT is supported by a Rubicon grant (680-50-1527) from the Netherlands Organisation for Scientific Research (NWO)

References (356)

  • P.J. Basser et al.

    MR diffusion tensor spectroscopy and imaging

    Biophys. J.

    (1994)
  • D. Benjamini et al.

    Use of marginal distributions constrained optimization (MADCO) for accelerated 2D MRI relaxometry and diffusometry

    J. Magn. Reson.

    (2016)
  • D. Benjamini et al.

    Magnetic resonance microdynamic imaging reveals distinct tissue microenvironments

    Neuroimage

    (2017)
  • D. Benjamini et al.

    White matter microstructure from nonparametric axon diameter distribution mapping

    Neuroimage

    (2016)
  • B. Bodini et al.

    Diffusion MRI in neurological disorders

  • K. Borkowski et al.

    A theoretical validation of the B-matrix spatial distribution approach to diffusion tensor imaging

    Magn. Reson. Imag.

    (2017)
  • M.D. Budde et al.

    Examining brain microstructure using structure tensor analysis of histological sections

    Neuroimage

    (2012)
  • L.M. Burcaw et al.

    Mesoscopic structure of neuronal tracts from time-dependent diffusion

    Neuroimage

    (2015)
  • P.T. Callaghan et al.

    Use of the second dimension in PGSE NMR studies of porous media

    Magn. Reson. Imag.

    (2003)
  • M. Catani et al.

    Short frontal lobe connections of the human brain

    Cortex J. Devoted Stud. Nerv. Syst. Behav.

    (2012)
  • M. Catani et al.

    Short parietal lobe connections of the human and monkey brain

    Cortex

    (2017)
  • H. Celik et al.

    Stabilization of the inverse Laplace transform of multiexponential decay through introduction of a second dimension

    J. Magn. Reson.

    (2013)
  • H. D'Arceuil et al.

    The effects of brain tissue decomposition on diffusion tensor imaging and tractography

    Neuroimage

    (2007)
  • H.E. D'Arceuil et al.

    An approach to high resolution diffusion tensor imaging in fixed primate brain

    Neuroimage

    (2007)
  • A. Daducci et al.

    Accelerated microstructure imaging via convex optimization (AMICO) from diffusion MRI data

    Neuroimage

    (2015)
  • S. De Santis et al.

    T1 relaxometry of crossing fibres in the human brain

    Neuroimage

    (2016)
  • S. De Santis et al.

    Including diffusion time dependence in the extra-axonal space improves in vivo estimates of axonal diameter and density in human white matter

    Neuroimage

    (2016)
  • B. Dhital et al.

    The absence of restricted water pool in brain white matter

    Neuroimage

    (2018)
  • M.D. Does et al.

    Multi-component T1 relaxation and magnetisation transfer in peripheral nerve

    Magn. Reson. Imag.

    (1998)
  • I. Drobnjak et al.

    Optimising time-varying gradient orientation for microstructure sensitivity in diffusion-weighted MR

    J. Magn. Reson.

    (2011)
  • I. Drobnjak et al.

    Optimizing gradient waveforms for microstructure sensitivity in diffusion-weighted MR

    J. Magn. Reson.

    (2010)
  • I. Drobnjak et al.

    The matrix formalism for generalised gradients with time-varying orientation in diffusion NMR

    J. Magn. Reson.

    (2011)
  • F. Du et al.

    Myelin and axon abnormalities in schizophrenia measured with magnetic resonance imaging techniques

    Biol. Psychiatr.

    (2013)
  • T. Duval et al.

    g-Ratio weighted imaging of the human spinal cord in vivo

    Neuroimage

    (2017)
  • J. Duyn

    MR susceptibility imaging

    J. Magn. Reson.

    (2013)
  • J.H. Duyn

    Studying brain microstructure with magnetic susceptibility contrast at high-field

    Neuroimage

    (2018)
  • F. Aboitiz et al.

    One hundred million years of interhemispheric communication: the history of the corpus callosum

    Braz. J. Med. Biol. Res.

    (2003)
  • A. Ahlgren et al.

    Quantification of microcirculatory parameters by joint analysis of flow-compensated and non-flow-compensated intravoxel incoherent motion (IVIM) data

    NMR Biomed.

    (2016)
  • S. Aja-Fernández et al.

    Statistical noise analysis in GRAPPA using a parametrized noncentral Chi approximation model

    Magn. Reson. Med.

    (2011 Apr)
  • D.C. Alexander et al.

    Image quality transfer via random forest regression: applications in diffusion MRI

  • D.C. Alexander et al.

    Imaging brain microstructure with diffusion MRI: practicality and applications

    NMR Biomed.

    (2017)
  • A.W. Anderson

    Measurement of fiber orientation distributions using high angular resolution diffusion imaging

    Magn. Reson. Med.

    (2005)
  • T.J. Andrews et al.

    Diffusion of myelin water

    Magn. Reson. Med.

    (2006)
  • D.S. Andrews et al.

    In vivo evidence of reduced integrity of the gray-white matter boundary in autism spectrum disorder

    Cerebr. Cortex

    (2017)
  • Y. Assaf et al.

    In vivo and in vitro bi-exponential diffusion of N-acetyl aspartate (NAA) in rat brain: a potential structural probe?

    NMR Biomed.

    (1998)
  • Y. Assaf et al.

    Structural information in neuronal tissue as revealed by q-space diffusion NMR spectroscopy of metabolites in bovine optic nerve

    NMR Biomed.

    (1999)
  • Y. Assaf et al.

    Assignment of the water slow-diffusing component in the central nervous system using q-space diffusion MRS: implications for fiber tract imaging

    Magn. Reson. Med.

    (2000)
  • Y. Assaf et al.

    High b-value q-space analyzed diffusion-weighted MRI: application to multiple sclerosis

    Magn. Reson. Med.

    (2002)
  • Y. Assaf et al.

    Axcaliber: a method for measuring axon diameter distribution from diffusion MRI

    Magn. Reson. Med.

    (2008)
  • J.C. Augustinack et al.

    Direct visualization of the perforant pathway in the human brain with ex vivo diffusion tensor imaging

    Front. Hum. Neurosci.

    (2010)
  • Cited by (126)

    View all citing articles on Scopus
    View full text