Synopsis
Typical diffusion MRI acquisitions
are blind to myelin, due to its short T2 time. Given that myelin comprises about 50% of the fiber volume, it is necessary to add a complementary myelin measure to better characterize white matter microstructure. Combining diffusion and myelin imaging sensitizes the MR measurement to the myelin g-ratio (a measure of myelin thickness), helping scientists gain novel insights into brain microstructure during development,
aging, and disease.Target audience
Scientists and clinicians with
a basic understanding of MRI physics, interested in modeling and non-invasive
characterization of white matter microstructure.
Purpose
Over the last ten years we
have seen tremendous advances in the field of quantitative magnetic resonance imaging,
enabling us to glean microstructural information on a scale that is orders of
magnitude smaller than the native MRI resolution. Advances in hardware and
pulse sequence design have enabled us to ask specific questions about the
distribution of axons and myelin in the brain, but the answers will have to
come from an interdisciplinary approach that combines multimodal imaging and
biophysical models of brain microstructure.
Let us imagine for a moment
that all white matter in the brain is comprised of parallel, circular,
concentric fibers, as shown in Fig. 1. Let us also imagine that the g-ratio
(defined as the ratio of the inner to the outer radius of the myelin sheath) is
constant for all fibers, and that there exists one qMRI measure specific to the
axon volume fraction (AVF), and another specific to the myelin volume fraction
(MVF). In this ideal world, there is a
very simple formula relating the MVF to the AVF (Fig. 1), and the key to that
relationship is the myelin g-ratio. For
the first part of this lecture, we will look at the ramifications of this
simplified model on the study of healthy and diseased brains. In the second part of the lecture we will
look at several ways in which the ideal model breaks down, and we will explore
what microstructural information about the real-world brain can be retained.
Methods
While diffusion models have
been used to describe the AVF, and there are a number of myelin models for characterizing
the MVF, it is only through combining these two that we can obtain a more
complete picture of the brain microstructure and the intricate relationship
between axon caliber and myelin thickness.
Typical diffusion MRI
acquisitions have relatively small signal contributions from myelin. This is
because the transverse relaxation time T2 of myelin water is short (10 - 30 ms [1]) and the echo time necessary to achieve sufficient
diffusion sensitization is long (∼100 ms). The lack of signal from the
myelin compartment in diffusion imaging means that estimation of the true
volume fractions of the other compartments is difficult.
Adding a
complementary myelin imaging technique brings us one step closer to properly
characterizing the brain microstructure.
Absolute myelin content
can be probed with techniques such as multicomponent T2 imaging [2], magnetization transfer [3] or T1 mapping [4].
Results
Several
papers have shown that diffusion and myelin imaging are complementary techniques
[5, 6],
where the former is more sensitive to the AVF and the latter is sensitive to
the MVF [7]. This suggests that combining the two can
distill important information about the white matter microstructure, in
particular about the relative myelin thickness, or the myelin g-ratio [7, 8]. In general, any study that measures some
quantitative parameters correlated with axon and myelin content will be
statistically sensitive to the myelin g-ratio [9].
Alexander et al. refer to a
combination of different MR contrasts, including diffusion and magnetization
transfer, as “quantitative stains” [10]. Recently,
several groups have combined diffusion and myelin imaging to characterize
myelin microstructure (See Fig. 2). For instance, diffusion
tensor imaging (DTI) and the magnetization transfer ratio (MTR) have been
combined to look at regional brain changes in myelination and structural
organization in early development [11]. More recently, myelin water fractions (MWF)
have complemented diffusion measures to observe the decrease in the g-ratio
during the early stages of myelination in preterm infants [12, 13]. An index sensitive to the myelin g-ratio has
also been reported in the brains [14] and spinal cords [15] of healthy adults,
suggesting that the g-ratio is relatively constant in adult white matter.
T1 mapping has also been used to complement
diffusion. Barazany et al. observed a negative correlation between the mean
axon diameter measured with AxCaliber [16] and the T1 relaxation time in white matter [17], an observation recently confirmed with histology [18]. While the
above result indicates that absolute myelin content is higher in regions with large
axons, this does not necessarily translate into greater relative myelin
thickness (lower g-ratio) in those regions.
As a matter of fact, super-axons found in the splenium of the corpus
callosum tend to have a higher g-ratio [19], and this was recently measured in vivo [9, 20] by
combining neurite orientation dispersion and density imaging (NODDI) [21] with quantitative magnetization transfer [22]. The ramifications of this approach are
particularly interesting in the context of multiple sclerosis (MS), where
variations in the g-ratio can be interpreted in terms of demyelination,
remyelination, and axonal loss [9, 23].
The field of MR tractometry [24] is another example of a multi-modal
approach that benefits from assigning quantitative MRI biomarkers to white
matter pathways derived from diffusion tractography.
The magnetization transfer
ratio is sensitive to myelin, and has been used to evaluate the level of
demyelination in MS [25]. Combining MTR and diffusion imaging
provides valuable information about the relationship between myelination and
fiber geometry [26, 27]. Looking at the magnetization transfer along fiber tracts
can help us identify the level of myelination of different fibers in the brain [28], as well as understand the patterns of
(de)myelination in normal-appearing white matter in MS patients and healthy
controls [29].
T1 tractometry has also been used to
evaluate myelination in white matter fibers.
A recent study found that each tract has a signature T1 value
that is consistent along its length for each subject [30].
While the T1 value
along a tract is nearly constant, the mean T1 value of a tract often
differs substantially from the T1 values of neighboring tracts in
the same hemisphere, and is consistent with myelination patterns during
development and aging. T1 tractometry in combination with the
CHARMED model has also succeeded in resolving both axonal and myelin properties
in the presence of multiple fiber populations within a voxel [31].
Discussion
Diffusion
and myelin imaging are complementary techniques, and combining them sensitizes
the measurement to the myelin g-ratio. The g-ratio framework in Fig. 1 holds
for many deviations from the ideal model.
Figures 3a) and 3b) illustrate several fiber arrangements for which the
framework holds, including non-parallel arbitrarily shaped fibers, uniform
thinning of the myelin sheath, and significant fiber loss. However, the framework hinges on assuming
concentric isomorphic shapes with a uniform g-ratio. Figure 3c) demonstrates a configuration with
non-uniform g-ratios, where the measurement will be biased towards the fiber
with larger caliber. Figure 3d) is an extreme
(and unrealistic) example with only two fibers, one with g ~ 0 and another with
g ~ 1. The average g-ratio in this
configuration is g = 0.5, but the g-ratio obtained with the formula from Fig. 1
is g = 0.7. This discrepancy should not
come as a surprise, as the g-ratio framework assumes homogeneity of the g-ratio
within the voxel, much like every other qMRI model that assigns a single number
(and not a distribution) to a voxel. Fortunately, histology shows that the
g-ratio is significantly more uniform within a voxel compared to AVF and MVF,
justifying the uniformity assumption [32]. However, to remove any ambiguity arising from
calling the computed metric an ‘average’ g-ratio, we recommend referring to it
as the ‘aggregate’ g-ratio.
While
the AVF and MVF vary significantly in the brain, the myelin g-ratio has a much
narrower dynamic range, which is theoretically predicted to be optimal for
values around 0.7 [33-35]. Deviations in the g-ratio have been observed
in several neurodegenerative diseases [36, 37], but to be sensitive to
these deviations we need the qMRI biomarker for AVF to not be influenced by the
MVF, and vice versa. However, this decoupling is impossible to achieve in a
realistic MR experiment, and as a result, the AVF will be a function of the
MVF. Hence, any imperfect calibration
between the MRI metric and the absolute MVF will produce artifactual g-ratio
trends, as shown in Fig. 4. This figure
raises the issue of which diffusion and myelin metrics, when combined, provide
the greatest specificity to the myelin g-ratio.
That is why, in addition to NODDI and magnetization transfer, there is
great value in exploring multi-modal imaging with other microstructural imaging
techniques, such as AxCaliber [16], ActiveAx [38], and MTV [39].
Conclusion
The promise of combining qMRI
measurements to characterize tissue is at the core of the newly emerging field
of in vivo
histology. Being able to
map the g-ratio non-invasively opens up a wide range of possibilities for the
study of white matter. Combined with measurement of the axon diameter
distribution, which is possible with techniques such as “AxCaliber” [16] and “ActiveAx” [38], the g-ratio will allow us to see a more
complete picture of white matter microstructure from imaging data. Thanks to
its potential for tracking microstructural changes during development, aging,
disease and treatment, multi-modal imaging has the promise to become an
invaluable tool in the in vivo histology
toolbox.
Acknowledgements
No acknowledgement found.References
1. Whittall, K.P., et al., In vivo measurement of T2 distributions and
water contents in normal human brain. Magn Reson Med, 1997. 37(1): p. 34-43.
2. MacKay,
A., et al., In vivo visualization of
myelin water in brain by magnetic resonance. Magn Reson Med, 1994. 31(6): p. 673-677.
3. Wolff,
S.D. and R.S. Balaban, Magnetization
transfer contrast (MTC) and tissue water proton relaxation in vivo. Magn
Reson Med, 1989. 10(1): p. 135-44.
4. Stuber,
C., et al., Myelin and iron concentration
in the human brain: a quantitative study of MRI contrast. Neuroimage, 2014.
93 Pt 1: p. 95-106.
5. Underhill,
H.R., C. Yuan, and V.L. Yarnykh, Direct
quantitative comparison between cross-relaxation imaging and diffusion tensor
imaging of the human brain at 3.0 T. Neuroimage, 2009. 47(4): p. 1568-78.
6. Madler,
B., et al., Is diffusion anisotropy an
accurate monitor of myelination? Correlation of multicomponent T2 relaxation and
diffusion tensor anisotropy in human brain. Magn Reson Imaging, 2008. 26(7): p. 874-888.
7. Stikov,
N., et al., Bound pool fractions
complement diffusion measures to describe white matter micro and
macrostructure. Neuroimage, 2011. 54(2):
p. 1112-21.
8. West,
K.L., et al., A revised model for
estimating g-ratio from MRI. Neuroimage, 2015.
9. Stikov,
N., et al., In vivo histology of the
myelin g-ratio with magnetic resonance imaging. Neuroimage, 2015. 118: p. 397-405.
10. Alexander,
A.L., et al., Characterization of
cerebral white matter properties using quantitative magnetic resonance imaging
stains. Brain Connect, 2011. 1(6):
p. 423-46.
11. Nossin-Manor,
R., et al., Quantitative MRI in the very
preterm brain: assessing tissue organization and myelination using
magnetization transfer, diffusion tensor and T1 imaging. Neuroimage,
2013. 64: p. 505-16.
12. Dean,
D.C., et al., Mapping an index of the
myelin g-ratio in infants using magnetic resonance imaging. Neuroimage,
2016.
13. Melbourne,
A., et al., Multi-modal measurement of
the myelin-to-axon diameter g-ratio in preterm-born neonates and adult
controls. Med Image Comput Comput Assist Interv, 2014. 17(Pt 2): p. 268-75.
14. Mohammadi,
S., et al., Whole-Brain In-vivo
Measurements of the Axonal G-Ratio in a Group of 37 Healthy Volunteers.
Front Neurosci, 2015. 9: p. 441.
15. Duval,
T., et al. In vivo mapping of myelin g-ratio
in the human spinal cord. in Proceedings
of the 22nd Annual Meeting of the International Society for Magnetic Resonance
in Medicine. 2015. Toronto.
16. Assaf,
Y., et al., AxCaliber: a method for
measuring axon diameter distribution from diffusion MRI. Magn Reson Med,
2008. 59(6): p. 1347-54.
17. Barazany,
D., D.K. Jones, and Y. Assaf. A combined
acquisition of T1 and AxCaliber can link between axon diameter and myelination in
Proceedings of the 22nd Annual Meeting of
the International Society for Magnetic Resonance in Medicine. 2014.
18. Harkins,
K.D., et al., The microstructural
correlates of t1 in white matter. Magnetic Resonance in Medicine, 2015.
19. Lamantia,
A.S. and P. Rakic, Cytological and
quantitative characteristics of four cerebral commissures in the rhesus monkey.
J Comp Neurol, 1990. 291(4): p.
520-37.
20. Stikov,
N., et al., Quantitative analysis of the
myelin g-ratio from electron microscopy images of the macaque corpus callosum.
Data Brief, 2015. 4: p. 368-73.
21. Zhang,
H., et al., NODDI: practical in vivo
neurite orientation dispersion and density imaging of the human brain.
Neuroimage, 2012. 61(4): p. 1000-16.
22. Levesque,
I.R., J.G. Sled, and G.B. Pike, Iterative
optimization method for design of quantitative magnetization transfer imaging
experiments. Magn Reson Med, 2011. 66(3):
p. 635-43.
23. Cercignani,
M., et al. Mapping the g-ratio within MS
lesions. in Proceedings of the 22nd
Annual Meeting of the International Society for Magnetic Resonance in Medicine.
2015. Toronto.
24. Bells,
S., et al., Tractometry–comprehensive
multi-modal quantitative assessment of white matter along specific tracts.
Proceedings of the 19th Annual Meeting of the International Society for
Magnetic Resonance in Medicine, 2011.
25. Schmierer,
K., et al., Magnetization transfer ratio
and myelin in postmortem multiple sclerosis brain. Ann Neurol, 2004. 56(3): p. 407-415.
26. De
Stefano, N., et al., Improving the
characterization of radiologically isolated syndrome suggestive of multiple
sclerosis. PLoS One, 2011. 6(4):
p. e19452.
27. Harrison,
D.M., et al., Tract-specific quantitative
MRI better correlates with disability than conventional MRI in multiple
sclerosis. J Neurol, 2013. 260(2):
p. 397-406.
28. Barazany,
D., et al. Characterizing white matter
pathways of the living rat brain by Tractometry. in Proceedings of the 22nd Annual Meeting of the International Society for
Magnetic Resonance in Medicine. 2014. Milan.
29. Stikov,
N., et al. Magnetization Transfer Ratio
Tractometry in Multiple Sclerosis. in Proceedings
of the 21st Annual Meeting of the International Society for Magnetic Resonance
in Medicine. 2013. Salt Lake City.
30. Yeatman,
J.D., B.A. Wandell, and A.A. Mezer, Lifespan
maturation and degeneration of human brain white matter. Nat Commun, 2014. 5: p. 4932.
31. De
Santis, S., et al. Resolving Myelin and
Axonal Properties within the Same Voxel in Presence of Crossing Fibers by
Combining Inversion Recovery and Diffusion Acquisitions. in Proceedings of the 22nd Annual Meeting of
the International Society for Magnetic Resonance in Medicine. 2014. Milan.
32. Stikov,
N., et al. In vivo measurement of the
myelin g-ratio with histological validation. in Proceedings of the 22nd Annual Meeting of the International Society for
Magnetic Resonance in Medicine. 2014. Milan.
33. Rushton,
W.A., A theory of the effects of fibre
size in medullated nerve. J Physiol, 1951. 115(1): p. 101-22.
34. Waxman,
S.G. and M.V. Bennett, Relative
conduction velocities of small myelinated and non-myelinated fibres in the
central nervous system. Nat New Biol, 1972. 238(85): p. 217-9.
35. Chomiak,
T. and B. Hu, What is the optimal value
of the g-ratio for myelinated fibers in the rat CNS? A theoretical approach.
PLoS One, 2009. 4(11): p. e7754.
36. Uranova,
N., et al., Electron microscopy of
oligodendroglia in severe mental illness. Brain Res Bull, 2001. 55(5): p. 597-610.
37. Albert,
M., et al., Extensive cortical
remyelination in patients with chronic multiple sclerosis. Brain Pathol,
2007. 17(2): p. 129-38.
38. Zhang,
H., et al., Axon diameter mapping in the
presence of orientation dispersion with diffusion MRI. Neuroimage, 2011. 56(3): p. 1301-15.
39. Mezer,
A., et al., Quantifying the local tissue
volume and composition in individual brains with magnetic resonance imaging.
Nat Med, 2013. 19(12): p. 1667-72.