Marco Palombo1
1Centre for Medical Image Computing (CMIC), University College London, United Kingdom
Synopsis
This lecture targets researchers and clinicians who are interested in using diffusion-weighted magnetic resonance spectroscopy (DW-MRS) of metabolites for brain microstructure characterisation. The audience will learn the basic mechanisms underpinning in-vivo metabolites DW-MRS, how to extract complex microstructural features characterising the morphology of specific cell-types (i.e. neurons and glia), together with some clinical and preclinical applications. Particular effort will be made to give intuitive insight and exiting perspectives on novel DW-MRS applications for brain microstructure quantification.
Target Audience
Researchers and clinicians who are interested in diffusion-weighted magnetic resonance spectroscopy (DW-MRS) of metabolites for brain microstructure characterisation and use or plan to use this technique to gain cell-type and intra-cellular specificity and to complement the information obtainable by other quantitative MRI methods.Objectives
The audience will learn the basic mechanisms underpinning in-vivo metabolites DW-MRS, how to extract complex microstructural features (i.e. cell body size, neurites length and size, branching order, spine density) characterising the morphology of specific cell-types (i.e. neurons and glia), together with some clinical and preclinical applications. Particular effort will be made to give intuitive insight and exiting perspectives on novel DW-MRS applications for brain microstructure quantification. In particular, the following aspects will be covered:
- Signal representation and biophysical modelling of metabolites DW-MRS signal;
- Realistic computational models and modern machine learning approaches for estimation of complex microstructural features from DW-MRS signal;
- Designing and optimising DW-MRS acquisitions to increase sensitivity to specific microstructural features;
- Preclinical applications and perspectives for clinical translation
Purpose
To provide the key concepts behind metabolite DW-MRS signal analysis for brain microstructure quantification and to show the audience various examples of possible applications.Overview
Diffusion-weighted MR spectroscopy (DW-MRS) enables the investigation of the intra-cellular environment through the measurement of the diffusion properties of several brain metabolites in vivo (1-4). The presence of restrictive or hindering frontiers (membranes, cytoskeleton, macromolecules, organelles etc.) may drastically influence the motion of brain metabolites and the consequent signal attenuation, as well as derived metrics such as the apparent diffusion coefficient (ADC) (5,6).
In contrast to water molecules, which are ubiquitous in biological tissue (inside any cell type and also in the extra-cellular space), metabolites are unique probes of the intra-cellular space, with typical extracellular concentrations at least 10
3-10
4 times lower than intra-cellular concentrations. Moreover, thanks to the preferential location of certain metabolites in specific cell types (e.g. N-AcetylAspartate, NAA, and Glutamate, Glu, preferentially in neurons while Myo-inositol, Ins, and Choline compounds, Cho, in glia), DW-MRS can help disentangling the effect of multiple pathological processes affecting tissue microstructure, and complement more sensitive but less specific methods such as diffusion tensor imaging (DTI).
Although very promising, DW-MRS also poses several challenges, which may limit the incorporation of this technique in standard MR protocols (1-4). Nevertheless, metabolite DW-MRS can provide unique and specific information on tissue properties, such as cytosol viscosity, tortuosity of the intra-cellular compartments, cell size and morphology, currently inaccessible with other MR techniques.
In this talk, we survey some recently proposed approaches to analyze brain metabolites DW-MRS data and estimate cell-specific microstructural features. Examples include:
- Non-Gaussianity
- Microscopic diffusion anisotropy
- Cell fibre caliber estimation
- Cell fibre microscopic orientation dispersion
- Cellular long-range microstructure: cell fiber segment length and number of embranchments.
Brain microstructure quantification using the diffusion of brain metabolites
We discuss the analysis of DW-MRS data using signal representation and biophysical and computational models (7) which link the measured diffusion sensitized echo signal attenuation and/or the ADC to cellular microstructural determinants, such as fibres diameter.
In the case of diffusion in brain tissue, the normalized echo signal attenuation is expected to be no longer a simple mono-exponential decay (typical of Gaussian diffusion) as a function of the diffusion weighting factor b, and the measured ADC is expected to be dependent on the diffusion time td. Pioneering works from Assaf and Cohen (8,9) characterized the non mono-exponential diffusion behavior of NAA in the brain by showing its bi- and tri-exponential diffusion decays within a large range of b values (up to 35000 s/mm2) and diffusion times (td up to 300 ms). More recently, Ligneul et al. (10) used b values up to 60000 s/mm2 in mouse brain to show a clear non mono-exponential signal attenuation for also other metabolites like tCho, Ins, tCr and Taurine (Tau) and extremely small or even nonexistent correlation between relaxation and diffusion, supporting the interpretation and modeling of these metabolites diffusion primarily based on geometry. Furthermore, Ingo et al. (11) recently showed non mono-exponential signal attenuation for NAA, tCr and tCho also in the human brain. These results suggest that brain metabolites diffusion may be restricted in cellular compartments, and consequently it may contain interesting information about the cellular microarchitecture.
Kroenke et al. (12) and Yablonskiy and Sukstanskii (13) proposed a first attempt to model NAA diffusion taking into account cellular microstructure by proposing a model of randomly oriented sticks (zero-radius cylinders). More recently, Palombo et al. (14) relaxed the zero-radius assumption and used a model of randomly oriented cylinders to characterize the diffusion of tCho, Ins, tCr, Taurine, Glu and NAA in mouse brain. This simple geometrical model provided estimates of cellular fibres size compatible with histology measurements and in agreement with the cell-type specific compartmentalization of metabolites. Shemesh et al. (15) used double diffusion encoding (DDE) measurements of NAA and Ins diffusion to measure neuronal and glial cell fibre microanisotropy and size. Using a similar model of randomly oriented cylinders, they estimated fibre sizes in good agreement with single diffusion encoding measurements (14). However, Vincent et al. (16) recently reported DDE measurements of also other metabolites (tCho, tCr, Glu) and showed that a more complex model (e.g. accounting for fibre branching and/or spines) is needed to fully characterize the DDE signal modulation.
Moving away from simple geometrical models, Palombo et al. (17-19) used numerical simulations and machine learning approaches to link brain metabolites diffusion properties to complex features of cellular morphology. They showed that metabolites signal attenuation at high b values and ADC time dependence can be used to quantify cell body size, cell fibre size and length, branching order, and spine density (17-19).
Some preclinical applications and perspectives for clinical translation of these techniques will be discussed. Acknowledgements
This work was supported by EPSRC grants EP/N018702/1.References
1. Nicolay K, Braun KP, Graaf RA, Dijkhuizen RM, & Kruiskamp MJ (2001) Diffusion NMR spectroscopy. Nmr Biomed 14(2):94-111.
2. Cao P & Wu EX (2017) In vivo diffusion MRS investigation of non-water molecules in biological tissues. Nmr Biomed 30(3).
3. Ronen I & Valette J (2015) Diffusion-weighted magnetic resonance spectroscopy. eMagRes 4:733–750.
4. Palombo M, Shemesh N, Ronen I, & Valette J (2018) Insights into brain microstructure from in vivo DW-MRS. Neuroimage 182:97-116.
5. Novikov DS, Fieremans E, Jespersen SN, & Kiselev VG (2019) Quantifying brain microstructure with diffusion MRI: Theory and parameter estimation. Nmr Biomed 32(4):e3998.
6. Valette J, Ligneul C, Marchadour C, Najac C, & Palombo M (2018) Brain Metabolite Diffusion from Ultra-Short to Ultra-Long Time Scales: What Do We Learn, Where Should We Go? Front Neurosci-Switz 12.
7. Novikov DS, Kiselev VG, & Jespersen SN (2018) On modeling. Magn Reson Med 79(6):3172-3193.
8. Assaf Y & Cohen Y (1998) Non-mono-exponential attenuation of water and N-acetyl aspartate signals due to diffusion in brain tissue. J Magn Reson 131(1):69-85.
9. Assaf Y & Cohen Y (1998) In vivo and in vitro bi-exponential diffusion of N-acetyl aspartate (NAA) in rat brain: a potential structural probe? Nmr Biomed 11(2):67-74.
10. Ligneul C, Palombo M, & Valette J (2016) Metabolite diffusion up to very high b in the mouse brain in vivo: Revisiting the potential correlation between relaxation and diffusion properties. Magn Reson Med.
11. Ingo C, Brink W, Ercan E, Webb AG, & Ronen I (2018) Studying neurons and glia non-invasively via anomalous subdiffusion of intracellular metabolites. Brain Structure & Function 223(8):3841-3854.
12. Kroenke CD, Ackerman JJ, & Yablonskiy DA (2004) On the nature of the NAA diffusion attenuated MR signal in the central nervous system. Magn Reson Med 52(5):1052-1059.
13. Yablonskiy DA & Sukstanskii AL (2010) Theoretical models of the diffusion weighted MR signal. Nmr Biomed23(7):661-681.
14. Palombo M, Ligneul C, & Valette J (2017) Modeling diffusion of intracellular metabolites in the mouse brain up to very high diffusion-weighting: Diffusion in long fibers (almost) accounts for non-monoexponential attenuation. Magnet Reson Med 77(1):343-350.
15. Shemesh N, et al. (2014) Metabolic properties in stroked rats revealed by relaxation-enhanced magnetic resonance spectroscopy at ultrahigh fields. Nat Commun 5:4958.
16. Vincent M, Palombo M, & Valette J (2020) Revisiting double diffusion encoding MRS in the mouse brain at 11.7T: Which microstructural features are we sensitive to? Neuroimage 207:116399.
17. Palombo M, et al. (2016) New paradigm to assess brain cell morphology by diffusion-weighted MR spectroscopy in vivo. Proc Natl Acad Sci U S A 113(24):6671-6676.
18. Palombo M, Ligneul C, Hernandez-Garzon E, & Valette J (2017) Can we detect the effect of spines and leaflets on the diffusion of brain intracellular metabolites? Neuroimage 182:283-293.
19. Ligneul C, et al. (2019) Diffusion-weighted magnetic resonance spectroscopy enables cell-specific monitoring of astrocyte reactivity in vivo. Neuroimage 191:457-469.