White Matter Model Research Articles

Discriminating Mild Traumatic Brain Injury and Posttraumatic Stress Disorder Using Latent Neuroimaging and Neuropsychological Profiles in Active-Duty Military Service Members.

Mild traumatic brain injury (mTBI) and posttraumatic stress disorder (PTSD) commonly occur among military Service Members and Veterans and have heterogenous, but also overlapping symptom presentations, which often complicate the diagnoses of underlying impairments and development of effective treatment plans. Thus, we sought to examine whether the combination of whole brain gray matter (GM) and white matter (WM) structural measures with neuropsychological performance can aid in the classification of military personnel with mTBI and PTSD. Active-Duty US Service Members ( n = 156; 87.8% male) with a history of mTBI, PTSD, combined mTBI+PTSD, or orthopedic injury completed a neuropsychological battery and T1- and diffusion-weighted structural neuroimaging. Cortical, subcortical, ventricular, and WM volumes and whole brain fractional anisotropy (FA), mean diffusivity (MD), radial diffusivity (RD), and axial diffusivity (AD) were calculated. Latent profile analyses were performed to determine how the GM and WM indicators, together with neuropsychological indicators, classified individuals. For both GM and WM, respectively, a 4-profile model was the best fit. The GM model identified greater ventricular volumes in Service Members with cognitive symptoms, including those with a diagnosis of mTBI, either alone or with PTSD. The WM model identified reduced FA and elevated RD in those with psychological symptoms, including those with PTSD or mTBI and comorbid PTSD. However, contrary to expectation, a global neural signature unique to those with comorbid mTBI and PTSD was not identified. The findings demonstrate that neuropsychological performance alone is more robust in differentiating Active-Duty Service Members with mTBI and PTSD, whereas global neuroimaging measures do not reliably differentiate between these groups.

Harmonic viscoelastic response of 3D histology-informed white matter model

White matter (WM) consists of bundles of long axons embedded in a glial matrix, which lead to anisotropic mechanical properties of brain tissue, and this complicates direct numerical simulations of WM viscoelastic response. The detailed axonal geometry contains scales that range from axonal diameter (microscale) to many diameters (mesoscale) imposing an additional challenge to numerical simulations. Here we describe the development of a 3D homogenization model for the central nervous system (CNS) that accounts for the anisotropy introduced by the axon/neuroglia composite, the axonal trace curvature, and the tissue dynamic response in the frequency domain. Homogenized models that allow the incorporation of all the above factors are important for accurately simulating the tissue's mechanical behavior, and this in turn is essential in interpreting non-invasive elastography measurements.Geometric and material parameters affect the material properties and thus the response of the brain tissue. More complex, orthotropic, or anisotropic material properties are to be considered as necessitated by the 3D tissue structure. An assembly of micro-scale 3D representative elemental volumes (REVs) is constructed, leading to an integrated mesoscale WM finite element model. Assemblies of microscopic REVs, with orientations based on geometrical reconstructions driven by confocal microscopy data are employed to form the elements of the WM model. Each REV carries local material properties based on a finite element model of biphasic (axon-glial matrix) unidirectional composite. The viscoelastic response of the microscopic REVs is extracted based on geometric information and fiber volume fractions calculated from the relative distance between the local elements and global axonal trace. The response of the WM tissue model is homogenized by averaging the shear moduli over the total volume (thus deriving effective properties) under realistic external loading conditions. Under harmonic shear loading, it is proven that that the effective transverse shear moduli are higher than the axial moduli when the axon moduli are higher than the glial. Methodologically, the process of using micro-scale 3D REVs to describe more complex axon geometries avoids the partition process in traditional composite finite element methods (based on partition of finite element grids) and constitutes a robust algorithm to automatically build a WM model based on available axonal trace information. Analytically, the model provides unmatched simulation flexibility and computational power as the position, orientation, and the magnitude of each tissue building block is calculated using real tissue data, as are the training and testing processes at each level of the multiscale WM tissue.

Joint independent component analysis for hypothesizing spatiotemporal relationships between longitudinal gray and white matter changes in preclinical Alzheimer's disease.

Characterizing relationships between gray matter (GM) and white matter (WM) in early Alzheimer's disease (AD) would improve understanding of how and when AD impacts the brain. However, modeling these relationships across brain regions and longitudinally remains a challenge. Thus, we propose extending joint independent component analysis (jICA) into spatiotemporal modeling of regional cortical thickness and WM bundle volumes leveraging multimodal MRI. We jointly characterize these GM and WM features in a normal aging (n=316) and an age- and sex-matched preclinical AD cohort (n=81) at each of two imaging sessions spaced three years apart, training on the normal aging population in cross-validation and interrogating the preclinical AD cohort. We find this joint model identifies reproducible, longitudinal changes in GM and WM between the two imaging sessions and that these changes are associated with preclinical AD and are plausible considering the literature. We compare this joint model to two focused models: (1) GM features at the first session and WM at the second and (2) vice versa. The joint model identifies components that correlate poorly with those from the focused models, suggesting the different models resolve different patterns. We find the strength of association with preclinical AD is improved in the GM to WM model, which supports the hypothesis that medial temporal and frontal thinning precedes volume loss in the uncinate fasciculus and inferior anterior-posterior association fibers. These results suggest that jICA effectively generates spatiotemporal hypotheses about GM and WM in preclinical AD, especially when specific intermodality relationships are considered a priori.

Decoding the microstructural properties of white matter using realistic models

Multi-echo gradient echo (ME-GRE) magnetic resonance signal evolution in white matter has a strong dependence on the orientation of myelinated axons with respect to the main static field. Although analytical solutions have been able to predict some of the white matter (WM) signal behaviour of the hollow cylinder model, it has been shown that realistic models of WM offer a better description of the signal behaviour observed.In this work, we present a pipeline to (i) generate realistic 2D WM models with their microstructure based on real axon morphology with adjustable fiber volume fraction (FVF) and g-ratio. We (ii) simulate their interaction with the static magnetic field to be able to simulate their MR signal. For the first time, we (iii) demonstrate that realistic 2D WM models can be used to simulate a MR signal that provides a good approximation of the signal obtained from a real 3D WM model derived from electron microscopy. We then (iv) demonstrate in silico that 2D WM models can be used to predict microstructural parameters in a robust way if ME-GRE multi-orientation data is available and the main fiber orientation in each pixel is known using DTI. A deep learning network was trained and characterized in its ability to recover the desired microstructural parameters such as FVF, g-ratio, free and bound water transverse relaxation and magnetic susceptibility. Finally, the network was trained to recover these micro-structural parameters from an ex vivo dataset acquired in 9 orientations with respect to the magnetic field and 12 echo times. We demonstrate that this is an overdetermined problem and that as few as 3 orientations can already provide comparable results for some of the decoded metrics.

Spherical deconvolution with tissue-specific response functions and multi-shell diffusion MRI to estimate multiple fiber orientation distributions (mFODs)

In diffusion MRI, spherical deconvolution approaches can estimate local white matter (WM) fiber orientation distributions (FOD) which can be used to produce fiber tractography reconstructions. The applicability of spherical deconvolution to gray matter (GM), however, is still limited, despite its critical role as start/endpoint of WM fiber pathways. The advent of multi-shell diffusion MRI data offers additional contrast to model the GM signal but, to date, only isotropic models have been applied to GM. Evidence from both histology and high-resolution diffusion MRI studies suggests a marked anisotropic character of the diffusion process in GM, which could be exploited to improve the description of the cortical organization. In this study, we investigated whether performing spherical deconvolution with tissue specific models of both WM and GM can improve the characterization of the latter while retaining state-of-the-art performances in WM. To this end, we developed a framework able to simultaneously accommodate multiple anisotropic response functions to estimate multiple, tissue-specific, fiber orientation distributions (mFODs). As proof of principle, we used the diffusion kurtosis imaging model to represent the WM signal, and the neurite orientation dispersion and density imaging (NODDI) model to represent the GM signal. The feasibility of the proposed approach is shown with numerical simulations and with data from the Human Connectome Project (HCP). The performance of our method is compared to the current state of the art, multi-shell constrained spherical deconvolution (MSCSD). The simulations show that with our new method we can accurately estimate a mixture of two FODs at SNR≥50. With HCP data, the proposed method was able to reconstruct both tangentially and radially oriented FODs in GM, and performed comparably well to MSCSD in computing FODs in WM. When performing fiber tractography, the trajectories reconstructed with mFODs reached the cortex with more spatial continuity and for a longer distance as compared to MSCSD and allowed to reconstruct short trajectories tangential to the cortical folding. In conclusion, we demonstrated that our proposed method allows to perform spherical deconvolution of multiple anisotropic response functions, specifically improving the performances of spherical deconvolution in GM tissue.

The effect of realistic geometries on the susceptibility-weighted MR signal in white matter.

PurposeTo investigate the effect of realistic microstructural geometry on the susceptibility‐weighted MR signal in white matter (WM), with application to demyelination.MethodsPrevious work has modeled susceptibility‐weighted signals under the assumption that axons are cylindrical. In this study, we explored the implications of this assumption by considering the effect of more realistic geometries. A three‐compartment WM model incorporating relevant properties based on the literature was used to predict the MR signal. Myelinated axons were modeled with several cross‐sectional geometries of increasing realism: nested circles, warped/elliptical circles, and measured axonal geometries from electron micrographs. Signal simulations from the different microstructural geometries were compared with measured signals from a cuprizone mouse model with varying degrees of demyelination.ResultsSimulation results suggest that axonal geometry affects the MR signal. Predictions with realistic models were significantly different compared with circular models under the same microstructural tissue properties, for simulations with and without diffusion.ConclusionThe geometry of axons affects the MR signal significantly. Literature estimates of myelin susceptibility, which are based on fitting biophysical models to the MR signal, are likely to be biased by the assumed geometry, as will any derived microstructural properties. Magn Reson Med 79:489–500, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Improved Automatic Segmentation of White Matter Hyperintensities in MRI Based on Multilevel Lesion Features.

Brain white matter hyperintensities (WMHs) are linked to increased risk of cerebrovascular and neurodegenerative diseases among the elderly. Consequently, detection and characterization of WMHs are of significant clinical importance. We propose a novel approach for WMH segmentation from multi-contrast MRI where both voxel-based and lesion-based information are used to improve overall performance in both volume-oriented and object-oriented metrics. Our segmentation method (AMOS-2D) consists of four stages following a "generate-and-test" approach: pre-processing, Gaussian white matter (WM) modelling, hierarchical multi-threshold WMH segmentation and object-based WMH filtering using support vector machines. Data from 28 subjects was used in this study covering a wide range of lesion loads. Volumetric T1-weighted images and 2D fluid attenuated inversion recovery (FLAIR) images were used as basis for the WM model and lesion masks defined manually in each subject by experts were used for training and evaluating the proposed method. The method obtained an average agreement (in terms of the Dice similarity coefficient, DSC) with experts equivalent to inter-expert agreement both in terms of WMH number (DSC=0.637 vs. 0.651) and volume (DSC=0.743 vs. 0.781). It allowed higher accuracy in detecting WMH compared to alternative methods tested and was further found to be insensitive to WMH lesion burden. Good agreement with expert annotations combined with stable performance largely independent of lesion burden suggests that AMOS-2D will be a valuable tool for fully automated WMH segmentation in patients with cerebrovascular and neurodegenerative pathologies.

Abnormalities in Diffusional Kurtosis Metrics Related to Head Impact Exposure in a Season of High School Varsity Football.

The purpose of this study was to determine whether the effects of cumulative head impacts during a season of high school football produce changes in diffusional kurtosis imaging (DKI) metrics in the absence of clinically diagnosed concussion. Subjects were recruited from a high school football team and were outfitted with the Head Impact Telemetry System (HITS) during all practices and games. Biomechanical head impact exposure metrics were calculated, including: total impacts, summed acceleration, and Risk Weighted Cumulative Exposure (RWE). Twenty-four players completed pre- and post-season magnetic resonance imaging, including DKI; players who experienced clinical concussion were excluded. Fourteen subjects completed pre- and post-season Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT). DKI-derived metrics included mean kurtosis (MK), axial kurtosis (K axial), and radial kurtosis (K radial), and white matter modeling (WMM) parameters included axonal water fraction, tortuosity of the extra-axonal space, extra-axonal diffusivity (De axial and radial), and intra-axonal diffusivity (Da). These metrics were used to determine the total number of abnormal voxels, defined as 2 standard deviations above or below the group mean. Linear regression analysis revealed a statistically significant relationship between RWE combined probability (RWECP) and MK. Secondary analysis of other DKI-derived and WMM metrics demonstrated statistically significant linear relationships with RWECP after covariate adjustment. These results were compared with the results of DTI-derived metrics from the same imaging sessions in this exact same cohort. Several of the DKI-derived scalars (Da, MK, K axial, and K radial) explained more variance, compared with RWECP, suggesting that DKI may be more sensitive to subconcussive head impacts. No significant relationships between DKI-derived metrics and ImPACT measures were found. It is important to note that the pathological implications of these metrics are not well understood. In summary, we demonstrate a single season of high school football can produce DKI measurable changes in the absence of clinically diagnosed concussion.

Differentiating T2 hyperintensity in neonatal white matter by two-compartment model of diffusional kurtosis imaging.

In conventional neonatal MRI, the T2 hyperintensity (T2h) in cerebral white matter (WM) at term-equivalent age due to immaturity or impairment is still difficult to identify. To clarify such issue, this study used the metrics derived from a two-compartment WM model of diffusional kurtosis imaging (WM-DKI), including intra-axonal, extra-axonal axial and radial diffusivities (Da, De,// and De,⊥), to compare WM differences between the simple T2h and normal control for both preterm and full-term neonates, and between simple T2h and complex T2h with hypoxic-ischemic encephalopathy (HIE). Results indicated that compared with control, the simple T2h showed significantly increased De,// and De,⊥, but no significant change in Da in multiple premyelination regions, indicative of expanding extra-axonal diffusion microenvironment; while myelinated regions showed no changes. However, compared with simple T2h, the complex T2h with HIE had decreased Da, increased De,⊥ in both premyelination and myelinated regions, indicative of both intra- and extra-axonal diffusion alterations. While diffusion tensor imaging (DTI) failed to distinguish simple T2h from complex T2h with HIE. In conclusion, superior to DTI-metrics, WM-DKI metrics showed more specificity for WM microstructural changes to distinguish simple T2h from complex T2h with HIE.

Differentiation between Subtypes of Primary Progressive Aphasia by Using Cortical Thickness and Diffusion-Tensor MR Imaging Measures.

To test a multimodal magnetic resonance (MR) imaging-based approach composed of cortical thickness and white matter (WM) damage metrics to discriminate between variants of primary progressive aphasia (PPA) that are nonfluent and/or agrammatic (NFVPPA) and semantic (SVPPA). This study was approved by the local ethics committees on human studies, and written informed consent from all patients was obtained before their enrollment. T1-weighted and diffusion-tensor (DT) MR images were obtained from 13 NFVPPA patients, 13 SVPPA patients, and 23 healthy control participants. Cortical thickness and DT MR imaging indices from the long-associative and interhemispheric WM tracts were obtained. A random forest (RF) analysis was used to identify the image features associated with each clinical syndrome. Individual patient classification was performed by using receiver operator characteristic curve analysis with cortical thickness, DT MR imaging, and a combination of the two modalities. RESULTS RF analysis showed that the best markers to differentiate the two PPA variants at an individual patient level among cortical thickness and DT MR imaging metrics were diffusivity abnormalities of the left inferior longitudinal and uncinate fasciculi and cortical thickness measures of the left temporal pole and inferior frontal gyrus. A combination of cortical thickness and DT MR imaging measures (the so-called gray-matter-and-WM model) was able to distinguish patients with NFVPPA and SVPPA with the following classification pattern: area under the curve, 0.91; accuracy, 0.89; sensitivity, 0.92; specificity, 0.85. Leave-one-out analysis demonstrated that the gray matter and WM model is more robust than the single MR modality models to distinguish PPA variants (accuracy was 0.86, 0.73, and 0.68 for the gray matter and WM model, the gray matter-only model, and the WM-only model, respectively). A combination of structural and DT MR imaging metrics may provide a quantitative procedure to distinguish NFVPPA and SVPPA patients at an individual patient level. The discrimination accuracies obtained suggest that the gray matter and WM model is potentially relevant for the differential diagnosis of the PPA variants in clinical practice.

Age related differences in reaction time components and diffusion properties of normal-appearing white matter in healthy adults

Deterioration of the white matter (WM) is viewed as the neural substrate of age differences in speed of information processing (reaction time, RT). However, the relationship between WM and RT components is rarely examined in healthy aging. We assessed the relationship between RT components derived from the Ratcliff diffusion model and micro-structural properties of normal-appearing WM (NAWM) in 90 healthy adults (age 18–82 years). We replicated all major extant findings pertaining to age differences in RT components and WM: lower drift rate, greater response conservativeness, longer non-decision time, lower fractional anisotropy (FA), greater mean (MD), axial (AD) and radial (RD) diffusivity were associated with advanced age. Age differences in anterior regions of the cerebral WM exceeded those in posterior regions. However, the only relationship between RT components and WM was the positive association between DR in the body of the corpus callosum and non-decision time. Thus, in healthy adults, age differences in NAWM diffusion properties are not a major contributor to age differences in RT components. Longitudinal studies with more precise and specific estimates of regional myelin content and evaluation of the contribution of age-related vascular risk factors are necessary to understand cerebral substrates of age-related cognitive slowing.

Histological correlation of diffusional kurtosis and white matter modeling metrics in cuprizone-induced corpus callosum demyelination.

The cuprizone mouse model is well established for studying the processes of both demyelination and remyelination in the corpus callosum, and it has been utilized together with diffusion tensor imaging (DTI) to investigate myelin and axonal pathology. Although some underlying morphological mechanisms contributing to the changes in diffusion tensor (DT) metrics have been identified, the understanding of specific associations between histology and diffusion measures remains limited. Diffusional kurtosis imaging (DKI) is an extension of DTI that provides metrics of diffusional non-Gaussianity, for which an associated white matter modeling (WMM) method has been developed. The main goal of the present study was to quantitatively assess the relationships between diffusion measures and histological measures in the mouse model of cuprizone-induced corpus callosum demyelination. The diffusional kurtosis (DK) and WMM metrics were found to provide additional information that enhances the sensitivity to detect the morphological heterogeneity in the chronic phase of the disease process in the rostral segment of the corpus callosum. Specifically, in the rostral segment, axonal water fraction (d = 2.6; p < 0.0001), radial kurtosis (d = 2.0; p = 0.001) and mean kurtosis (d = 1.5; p = 0.005) showed the most sensitivity between groups with respect to yielding statistically significant p values and high Cohen's d values. These results demonstrate the ability of DK and WMM metrics to detect white mater changes and inflammatory processes associated with cuprizone-induced demyelination. They also validate, in part, the application of these new WMM metrics for studying neurological diseases, as well as helping to elucidate their biophysical meaning.

Aging in deep gray matter and white matter revealed by diffusional kurtosis imaging

Diffusion tensor imaging has already been extensively used to probe microstructural alterations in white matter tracts, and scarcely, in deep gray matter. However, results in literature regarding age-related degenerative mechanisms in white matter tracts and parametric changes in the putamen are inconsistent. Diffusional kurtosis imaging is a mathematical extension of diffusion tensor imaging, which could more comprehensively mirror microstructure, particularly in isotropic tissues such as gray matter. In this study, we used the diffusional kurtosis imaging method and a white-matter model that provided metrics of explicit neurobiological interpretations in healthy participants (58 in total, aged from 25 to 84 years). Tract-based whole-brain analyses and regions-of-interest (anterior and posterior limbs of the internal capsule, cerebral peduncle, fornix, genu and splenium of corpus callosum, globus pallidus, substantia nigra, red nucleus, putamen, caudate nucleus, and thalamus) analyses were performed to examine parametric differences across regions and correlations with age. In white matter tracts, evidence was found supportive for anterior–posterior gradient and not completely supportive for retrogenesis theory. Age-related degenerations appeared to be broadly driven by axonal loss. Demyelination may also be a major driving mechanism, although confined to the anterior brain. In terms of deep gray matter, higher mean kurtosis and fractional anisotropy in the globus pallidus, substantia nigra, and red nucleus reflected higher microstructural complexity and directionality compared with the putamen, caudate nucleus, and thalamus. In particular, the unique age-related positive correlations for fractional anisotropy, mean kurtosis, and radial kurtosis in the putamen opposite to those in other regions call for further investigation of exact underlying mechanisms. In summary, the results suggested that diffusional kurtosis can provide measurements in a new dimension that were complementary to diffusivity metrics. Kurtosis together with diffusivity can more comprehensively characterize microstructural compositions and age-related changes than diffusivity alone. Combined with proper model, it may also assist in providing neurobiological interpretations of the identified alterations.

Effects of white matter microstructure on phase and susceptibility maps

PurposeTo investigate the effects on quantitative susceptibility mapping (QSM) and susceptibility tensor imaging (STI) of the frequency variation produced by the microstructure of white matter (WM).MethodsThe frequency offsets in a WM tissue sample that are not explained by the effect of bulk isotropic or anisotropic magnetic susceptibility, but rather result from the local microstructure, were characterized for the first time. QSM and STI were then applied to simulated frequency maps that were calculated using a digitized whole-brain, WM model formed from anatomical and diffusion tensor imaging data acquired from a volunteer. In this model, the magnitudes of the frequency contributions due to anisotropy and microstructure were derived from the results of the tissue experiments.ResultsThe simulations suggest that the frequency contribution of microstructure is much larger than that due to bulk effects of anisotropic magnetic susceptibility. In QSM, the microstructure contribution introduced artificial WM heterogeneity. For the STI processing, the microstructure contribution caused the susceptibility anisotropy to be significantly overestimated.ConclusionMicrostructure-related phase offsets in WM yield artifacts in the calculated susceptibility maps. If susceptibility mapping is to become a robust MRI technique, further research should be carried out to reduce the confounding effects of microstructure-related frequency contributions. Magn Reson Med 73:1258–1269, 2015. © 2014 Wiley Periodicals, Inc.