Detection of microscopic anisotropy in gray matter and in a novel tissue phantom using double Pulsed Gradient Spin Echo MR

https://doi.org/10.1016/j.jmr.2007.07.003Get rights and content

Abstract

A double Pulsed Gradient Spin Echo (d-PGSE) MR experiment was used to measure and assess the degree of local diffusion anisotropy in brain gray matter, and in a novel “gray matter” phantom that consists of randomly oriented tubes filled with water. In both samples, isotropic diffusion was observed at a macroscopic scale while anisotropic diffusion was observed at a microscopic scale, however, the nature of the resulting echo attenuation profiles were qualitatively different. Gray matter, which contains multiple cell types and fibers, exhibits a more complicated echo attenuation profile than the phantom. Since microscopic anisotropy was observed in both samples in the low q regime comparable to that achievable in clinical scanner, it may offer a new potential contrast mechanism for characterizing gray matter microstructure in medical and biological applications.

Introduction

Diffusion weighted imaging (DWI) is commonly used to characterize normal brain structure and brain pathology [1]. Diffusion tensor MRI (DTI) measurements, in particular, reveal significant differences in the macroscopic structure of gray and white matter. Using DTI, white matter typically appears to be anisotropic and gray matter appears isotropic [2]. This difference can be attributed to differences in the anatomical and architectural organization of these tissues.

White matter consists of highly ordered bundles at the molecular, microscopic, and macroscopic length scales with its parallel structure often exceeding the MRI voxel length scale (Fig. 1d–f). Its macroscopic anisotropic diffusion profile can be adequately characterized by an anisotropic apparent diffusion tensor [3].

Gray matter, on the other hand, has multiple cell types such as neuronal cell bodies and processes, randomly oriented axons, dendritic fibers, oligodendrocytes, extracellular matrices, etc. [4]. While each axonal projection and dendritic fiber has a particular orientation, which may exhibit an anisotropic diffusion profile on a microscopic scale, these neural processes are randomly oriented at the macroscopic voxel length scale (Fig. 1a and b). Thus, the isotropic diffusion tensor that is measured in gray matter is likely the result of powder averaging of the local anisotropic diffusion profiles over a myriad of microdomains [2] (see Fig. 1c).

Although, for macroscopically anisotropic materials like white matter the use of Pulsed Gradient Spin Echo [5], [6] based sequences is sufficient to reveal their anisotropy, this sequence fails to detect anisotropy in microscopically anisotropic but macroscopically isotropic materials. For the later case, the use of multi-gradient orientation techniques has been proposed.

In this work the double Pulsed Gradient Spin Echo (d-PGSE) experiment [7], [8], [9] is used to detect or discover whether gray matter exhibits microscopic diffusion anisotropy. The d-PGSE sequence (Fig. 2) and its two-dimensional variants [10] are already well-established techniques in non-medical applications to characterize local anisotropy of macroscopically isotropic materials, such as liquid crystals [7], [11] prolate yeast cells [8] and plants [12].

The d-PGSE sequence consists of two single-PGSE blocks, which are concatenated. The resulting spins from the first PGSE block become the population of spins interrogated by the second PGSE block. Because the resulting echoes depend on the spin evolution in both encoding periods, these contain information about the spins’ diffusion histories during both PGSE blocks.

To assess the presence of microscopic diffusion anisotropy, one compares two d-PGSE experiments in which diffusion sensitizing gradients are applied in the same and in orthogonal directions. For microscopically isotropic materials, regardless of the diffusion gradient encoding directions, the resulting echo attenuations all superimpose. However, in the case of materials that exhibit local anisotropy, the resulting curves observed from the collinear and orthogonal diffusion gradient encoding directions do not superimpose. Consequently, a difference between these curves indicates microscopic anisotropy.

To explore the origin of gray matter anisotropy, we also constructed a “gray matter” phantom that is macroscopically isotropic and microscopically anisotropic. The phantom is designed to be stable, so it can also be used as a diffusion standard for calibrating the d-PGSE sequences and NMR hardware. Furthermore, the phantom has a simple geometry so that the displacement history of spins can be mathematically modeled.

Section snippets

Experimental design

The double-PGSE sequence was applied in nine different combinations of gradient directions between the two pairs of gradient pulses (PGSE blocks). Three collinear directions: X_X, Y_Y and Z_Z; and six orthogonal combinations: X_Y, Y_X, X_Z, Z_X, Y_Z and Z_Y, were used. Note that combinations like X_Y and Y_X should yield the same attenuation profile, so differences between them can be used to assess system software and hardware performance. The second echo time was chosen to be different than

Single-PGSE

Fig. 3a–c shows PGSE signal attenuations for the “gray matter” phantom, fixed cortical tissue and fixed white matter. The cortical tissue (Fig. 3a) echo attenuations fall on the same curve, regardless of the diffusion gradient direction, indicating macroscopic isotropy. In the “gray matter” phantom (Fig. 3b) slight macroscopic anisotropy can be observed along the Z direction. This small deviation from the X and Y directions might result from the sedimentation process during phantom preparation.

Conclusions

This experiment provides preliminary support to the hypothesis that gray matter consists of microscopic anisotropic domains that are detectable by using a multi-dimensional gradient technique such as the d-PGSE.

Local anisotropy of a “gray matter” phantom was also detected using the d-PGSE technique. For the “gray matter” phantom, simulations agree well with the experimental data when short Δ were used. In this regime the approximation of a Gaussian displacement distribution function still holds

Acknowledgments

The authors thank Dr. H.D. Morris and Dr. M.J. Lizak for helping with all technical aspects of the experiments and Dr. R.E. Rycyna from Bruker Biospin for writing the NMR pulse sequences. We also thank the people who were involved in providing biological specimens: Mr. R.R. Clevenger, Mr. T.J. Hunt, Ms. G.J. Zywicke, Mr. A.D. Zetts, Mrs. K. Keeran, Mr. S.M. Kozlov and Mr. K.R. Jeffries, from LAMS, NHLBI, for supplying fixed pig spinal cord tissue, and Dr. M.A. Eckhaus and Dr. M.F. Starost, from

References (25)

  • Y. Cheng et al.

    Multiple scattering by NMR

    J. Am. Chem. Soc.

    (1999)
  • P.P. Mitra

    Multiple wave-vector extension of the NMR pulsed-field-gradient spin-echo diffusion measurement

    Phys. Rev. B

    (1995)
  • Cited by (100)

    • In vivo Correlation Tensor MRI reveals microscopic kurtosis in the human brain on a clinical 3T scanner

      2022, NeuroImage
      Citation Excerpt :

      Since then, DDE in the long mixing time regime has been extensively adopted to provide different parameters associated with microscopic diffusion anisotropy (a quantity related to Kaniso, see methods) which reports on the anisotropy of the tissue at the pore or microdomain length scale independently to confounding orientation dispersion effects (e.g. Jespersen et al., 2013; Lasič et al., 2014; Shemesh et al., 2016; Ianuş et al., 2018; Henriques et al., 2020). Microscopic diffusion anisotropy has been investigated using DDE in preclinical systems in both white and grey matter, e.g.: in fixed grey matter (Komlosh et al., 2007), pig spinal cord (Komlosh et al., 2008), pig optic nerve and cortical grey matter, with phantoms mimicking their respective microstructures (Shemesh et al., 2010b, 2011; Shemesh and Cohen, 2011; Shemesh et al., 2012a), in both in vivo and ex vivo rat brain (Shemesh et al., 2012b; Ianuş et al., 2018; Kerkelä et al., 2019), in ex vivo monkey brain (Jespersen et al., 2013), in rat spinal cord injury (Budde et al., 2017), and in ex vivo rat spinal cord (Shemesh, 2018). In clinical systems DDE microscopic diffusion anisotropy estimates have been reported in pig spinal cord (Lawrenz and Finsterbusch, 2011), and successfully mapped in the in vivo healthy human brain white matter (Lawrenz and Finsterbusch, 2013, 2015), cortical grey matter (Lawrenz and Finsterbusch, 2019), in the study of ageing (Lawrenz et al., 2016), and in clinical applications, where it showed increased specificity compared to DTI-derived FA in multiple sclerosis lesions (Yang et al., 2018) and has been investigated in Parkinson's disease (Kamiya et al., 2020).

    View all citing articles on Scopus
    View full text