Training a neural network for Gibbs and noise removal in diffusion MRI.

PURPOSE: To develop and evaluate a neural network-based method for Gibbs artifact and noise removal. METHODS: A convolutional neural network (CNN) was designed for artifact removal in diffusion-weighted imaging data. Two implementations were considered: one for magnitude images and one for complex images. Both models were based on the same encoder-decoder structure and were trained by simulating MRI acquisitions on synthetic non-MRI images. RESULTS: Both machine learning methods were able to mitigate artifacts in diffusion-weighted images and diffusion parameter maps. The CNN for complex images was also able to reduce artifacts in partial Fourier acquisitions. CONCLUSIONS: The proposed CNNs extend the ability of artifact correction in diffusion MRI. The machine learning method described here can be applied on each imaging slice independently, allowing it to be used flexibly in clinical applications.

A time-dependent diffusion MRI signature of axon caliber variations and beading.

MRI provides a unique non-invasive window into the brain, yet is limited to millimeter resolution, orders of magnitude coarser than cell dimensions. Here, we show that diffusion MRI is sensitive to the micrometer-scale variations in axon caliber or pathological beading, by identifying a signature power-law diffusion time-dependence of the along-fiber diffusion coefficient. We observe this signature in human brain white matter and identify its origins by Monte Carlo simulations in realistic substrates from 3-dimensional electron microscopy of mouse corpus callosum. Simulations reveal that the time-dependence originates from axon caliber variation, rather than from mitochondria or axonal undulations. We report a decreased amplitude of time-dependence in multiple sclerosis lesions, illustrating the potential sensitivity of our method to axonal beading in a plethora of neurodegenerative disorders. This specificity to microstructure offers an exciting possibility of bridging across scales to image cellular-level pathology with a clinically feasible MRI technique.

In vivo observation and biophysical interpretation of time-dependent diffusion in human cortical gray matter.

The dependence of the diffusion MRI signal on the diffusion time t is a hallmark of tissue microstructure at the scale of the diffusion length. Here we measure the time-dependence of the mean diffusivity D(t) and mean kurtosis K(t) in cortical gray matter and in 25 â€‹gray matter sub-regions, in 10 healthy subjects. Significant diffusivity and kurtosis time-dependence is observed for t=21.2-100 â€‹ms, and is characterized by a power-law tail ∼t-ϑ with dynamical exponent ϑ. To interpret our measurements, we systematize the relevant scenarios and mechanisms for diffusion time-dependence in the brain. Using the effective medium theory formalism, we derive an exact relation between the power-law tails in D(t) and K(t). The estimated dynamical exponent ϑ≃1/2 in both D(t) and K(t) is consistent with one-dimensional diffusion in the presence of randomly positioned restrictions along neurites. We analyze the short-range disordered statistics of synapses on axon collaterals in the cortex, and perform one-dimensional Monte Carlo simulations of diffusion restricted by permeable barriers with a similar randomness in their placement, to confirm the ϑ=1/2 exponent. In contrast, the Kärger model of exchange is less consistent with the data since it does not capture the diffusivity time-dependence, and the estimated exchange time from K(t) falls below our measured t-range. Although we cannot exclude exchange as a contributing factor, we argue that structural disorder along neurites is mainly responsible for the observed time-dependence of diffusivity and kurtosis. Our observation and theoretical interpretation of the t-1/2 tail in D(t) and K(t) altogether establish the sensitivity of a macroscopic MRI signal to micrometer-scale structural heterogeneities along neurites in human gray matter in vivo.

Inner SPACE: 400-Micron Isotropic Resolution MRI of the Human Brain.

OBJECTIVES: Clinically relevant neuroanatomy is challenging to teach, learn and remember since many functionally important structures are visualized best using histology stains from serial 2D planar sections of the brain. In clinical patients, the locations of specific structures then must be inferred from spatial position and surface anatomy. A 3D MRI dataset of neuroanatomy has several advantages including simultaneous multi-planar visualization in the same brain, direct end-user manipulation of the data and image contrast identical to clinical MRI. We created 3D MRI datasets of the postmortem brain with high spatial and contrast resolution for simultaneous multi-planar visualization of complex neuroanatomy. MATERIALS AND METHODS: Whole human brains (N = 6) were immersion-fixed in 4% formaldehyde for 4 weeks, then washed continuously in water for 48 h. The brains were imaged on a clinical 3-T MRI scanner with a 64-channel head and neck coil using a 3D T2-weighted sequence with 400-micron isotropic resolution (voxel = 0.064 mm3; time = 7 h). Besides resolution, this sequence has multiple adjustments to improve contrast compared to a clinical protocol, including 93% reduced turbo factor and 77% reduced effective echo time. RESULTS: This MRI microscopy protocol provided excellent contrast resolution of small nuclei and internal myelinated pathways within the basal ganglia, thalamus, brainstem, and cerebellum. Contrast was sufficient to visualize the presence and variation of horizontal layers in the cerebral cortex. 3D isotropic resolution datasets facilitated simultaneous multi-planar visualization and efficient production of specific tailored oblique image orientations to improve understanding of complex neuroanatomy. CONCLUSION: We created an unlabeled high-resolution digital 3D MRI dataset of neuroanatomy as an online resource for readers to download, manipulate, annotate and use for clinical practice, research, and teaching that is complementary to traditional histology-based atlases. Digital MRI contrast is quantifiable, reproducible across brains and could help validate novel MRI strategies for in vivo structure visualization.

Observation of structural universality in disordered systems using bulk diffusion measurement.

We report on an experimental observation of classical diffusion distinguishing between structural universality classes of disordered systems in one dimension. Samples of hyperuniform and short-range disorder were designed, characterized by the statistics of the placement of micrometer-thin parallel permeable barriers, and the time-dependent diffusion coefficient was measured by NMR methods over three orders of magnitude in time. The relation between the structural exponent, characterizing disorder universality class, and the dynamical exponent of the diffusion coefficient is experimentally verified. The experimentally established relation between structure and transport exemplifies the hierarchical nature of structural complexity-dynamics are mainly determined by the universality class, whereas microscopic parameters affect the nonuniversal coefficients. These results open the way for noninvasive characterization of structural correlations in porous media, complex materials, and biological tissues via a bulk diffusion measurement.

The Behavior of Water in Collagen and Hydroxyapatite Sites of Cortical Bone: Fracture, Mechanical Wear, and Load Bearing Studies.

The mechanical properties of cortical bone, which is largely comprised of collagen, hydroxyapatite, and water, are known to hinge on hydration. Recently, the characteristics of water in bone have drawn attention as potential markers of bone quality. We report on the dynamics, diffusion, population, and exchange of water in cortical bone by NMR relaxation and diffusion methodologies. Relaxation measurements over timescales ranging from 0.001 to 4.2 s reveal two distinguishable water environments. Systematic exposure to ethylenediaminetetraacetic acid or collagenase reveals one peak in our 2D relaxation map belonging to water present in the hydroxyapatite rich environment, and a second peak with shorter relaxation times arising from a collagen rich site. Diffusion-T2 measurements allowed for direct measurement of the diffusion coefficient of water in all observable reservoirs. Further, deuterium relaxation methods were applied to study cortical bone under an applied force, following mechanical wear or fracture. The tumbling correlation times of water reduce in all three cases, indicating that water dynamics may be used as a probe of bone quality. Lastly, changes in the relative populations and correlation times of water in bone under an applied force suggest that load bearing occurs largely in the collagen rich environment and is reversible.