Publications by Year: 2010

BACKGROUND: While the neuroanatomical underpinnings of the functional brain disconnectivity observed in patients with schizophrenia (SZ) remain elusive, white matter fiber bundles of the brain are a likely candidate, given that they represent the infrastructure for long-distance neural communication. METHODS: This study investigated for diffusion abnormalities in 19 patients with chronic SZ, relative to 19 matched control subjects, across tractography-defined segments of the corpus callosum. Diffusion-weighted images were acquired with 51 noncollinear gradients on a 3T scanner (1.7 mm isotropic voxels). The corpus callosum was extracted by means of whole-brain tractography and automated fiber clustering and was parcelled into six segments on the basis of fiber trajectories. The diffusion indexes of fractional anisotropy (FA) and mode were calculated for each segment. RESULTS: Relative to the healthy control subjects, the SZ patients exhibited mode increases in the parietal fibers, suggesting a relative absence of crossing fibers. Schizophrenia patients also exhibited FA reductions in the frontal fibers, which were underpinned by increases in radial diffusivity, consistent with myelin abnormalities. Significant correlations were observed between patients’ degree of reality distortion and their FA and radial diffusivity, such that the most severely psychotic patients were the least abnormal in terms of their frontal fiber diffusivity. CONCLUSIONS: The SZ patients exhibited a variety of diffusion abnormalities in the corpus callosum, which were related to the severity of their psychotic symptoms. To the extent that diffusion abnormalities influence axonal transmission velocities, these results provide support for those theories that emphasize neural timing abnormalities in the etiology of schizophrenia.

Aging is associated with a variety of structural and pathological brain changes. Memory, or the ability to store and retrieve information, declines significantly during aging. In order to characterize the brain micro-structural correlates of memory performance, 52 healthy subjects, age 25-82 years, completed a computerized non-verbal memory test and were scanned using magnetic resonance diffusion tensor imaging. Partial correlation was conducted between DTI parameters and memory performance (accuracy and reaction time (RT) for different learning stages) controlling for age. A similar correlation pattern was found for apparent diffusion coefficient (ADC), FA, and radial and axial diffusivities, but correlations between ADC and memory performance were the most informative and are therefore reported here. While ADC was correlated with accuracy mainly in temporal and frontal cortical regions, it was correlated with RT in temporal and frontal white matter pathways, including the inferior longitudinal fasciculus and uncinate fasciculus. The task was repeated four times, performance in the first repetition was correlated with ADC in frontal white matter and in the fourth repetition with ADC in temporal gray matter structures mainly the parahippocapus and in the middle temporal gyrus. The localization of the correlations of ADC with the different task parameters is in line with previous studies. Thus, inter-subject variability in memory performance and tissue morphology, as expressed by diffusion tensor magnetic resonance imaging, can be used to relate tissue microstructures with cognitive performance, and to provide information to corroborate other functional localization techniques.

NMR observable nuclei undergoing restricted diffusion within confining pores are important reporters for microstructural features of porous media including, inter-alia, biological tissues, emulsions and rocks. Diffusion NMR, and especially the single-pulsed field gradient (s-PFG) methodology, is one of the most important noninvasive tools for studying such opaque samples, enabling extraction of important microstructural information from diffusion-diffraction phenomena. However, when the pores are not monodisperse and are characterized by a size distribution, the diffusion-diffraction patterns disappear from the signal decay, and the relevant microstructural information is mostly lost. A recent theoretical study predicted that the diffusion-diffraction patterns in double-PFG (d-PFG) experiments have unique characteristics, such as zero-crossings, that make them more robust with respect to size distributions. In this study, we theoretically compared the signal decay arising from diffusion in isolated cylindrical pores characterized by lognormal size distributions in both s-PFG and d-PFG methodologies using a recently presented general framework for treating diffusion in NMR experiments. We showed the gradual loss of diffusion-diffraction patterns in broadening size distributions in s-PFG and the robustness of the zero-crossings in d-PFG even for very large standard deviations of the size distribution. We then performed s-PFG and d-PFG experiments on well-controlled size distribution phantoms in which the ground-truth is well-known a priori. We showed that the microstructural information, as manifested in the diffusion-diffraction patterns, is lost in the s-PFG experiments, whereas in d-PFG experiments the zero-crossings of the signal persist from which relevant microstructural information can be extracted. This study provides a proof of concept that d-PFG may be useful in obtaining important microstructural features in samples characterized by size distributions.

One of the hallmarks of diffusion NMR and MRI is its ability to utilize restricted diffusion to probe compartments much smaller than the excited volume or the MRI voxel, respectively, and to extract microstructural information from them. Single-pulsed field gradient (s-PFG) MR methodologies have been employed with great success to probe microstructures in various disciplines, ranging from chemistry to neuroscience. However, s-PFG MR also suffers from inherent shortcomings, especially when specimens are characterized by orientation or size distributions: in such cases, the microstructural information available from s-PFG experiments is limited or lost. Double-pulsed field gradient (d-PFG) MR methodology, an extension of s-PFG MR, has attracted attention owing to recent theoretical studies predicting that it can overcome certain inherent limitations of s-PFG MR. In this review, we survey the microstructural features that can be obtained from conventional s-PFG methods in the different q regimes, and highlight its limitations. The experimental aspects of d-PFG methodology are then presented, together with an overview of its theoretical underpinnings and a general framework for relating the MR signal decay and material microstructure, affording new microstructural parameters. We then discuss recent studies that have validated the theory using phantoms in which the ground truth is well known a priori, a crucial step prior to the application of d-PFG methodology in neuronal tissue. The experimental findings are in excellent agreement with the theoretical predictions and reveal, inter alia, zero-crossings of the signal decay, robustness towards size distributions and angular dependences of the signal decay from which accurate microstructural parameters, such as compartment size and even shape, can be extracted. Finally, we show some initial findings in d-PFG MR imaging. This review lays the foundation for future studies, in which accurate and novel microstructural information could be extracted from complex biological specimens, eventually leading to new forms of contrast in MRI.

Noninvasive characterization of pore size and shape in opaque porous media is a formidable challenge. NMR diffusion-diffraction patterns were found to be exceptionally useful for obtaining such morphological features, but only when pores are monodisperse and coherently placed. When locally anisotropic pores are randomly oriented, conventional diffusion NMR methods fail. Here, we present a simple, direct, and general approach to obtain both compartment size and shape even in such settings and even when pores are characterized by internal field gradients. Using controlled porous media, we show that the bipolar-double-pulsed-field-gradient (bp-d-PFG) methodology yields diffusion-diffraction patterns from which pore size can be directly obtained. Moreover, we show that pore shape, which cannot be obtained by conventional methods, can be directly inferred from the modulation of the signal in angular bp-d-PFG experiments. This new methodology significantly broadens the types of porous media that can be studied using noninvasive diffusion-diffraction NMR.

We describe a technique that uses tractography to drive the local fiber model estimation. Existing techniques use independent estimation at each voxel so there is no running knowledge of confidence in the estimated model fit. We formulate fiber tracking as recursive estimation: at each step of tracing the fiber, the current estimate is guided by those previous. To do this we perform tractography within a filter framework and use a discrete mixture of Gaussian tensors to model the signal. Starting from a seed point, each fiber is traced to its termination using an unscented Kalman filter to simultaneously fit the local model to the signal and propagate in the most consistent direction. Despite the presence of noise and uncertainty, this provides a causal estimate of the local structure at each point along the fiber. Using two- and three-fiber models we demonstrate in synthetic experiments that this approach significantly improves the angular resolution at crossings and branchings. In vivo experiments confirm the ability to trace through regions known to contain such crossing and branching while providing inherent path regularization.

We propose a novel probabilistic framework to merge information from DWI tractography and resting-state fMRI correlations. In particular, we model the interaction of latent anatomical and functional connectivity templates between brain regions and present an intuitive extension to population studies. We employ a mean-field approximation to fit the new model to the data. The resulting algorithm identifies differences in latent connectivity between the groups. We demonstrate our method on a study of normal controls and schizophrenia patients.

We introduce a fibre tractography framework based on a particle filter which estimates a local geometrical model of the underlying white matter tract, formulated as a ’streamline flow’ using generalized helicoids. The method is not dependent on the diffusion model, and is applicable to diffusion tensor (DT) data as well as to high angular resolution reconstructions. The geometrical model allows for a robust inference of local tract geometry, which, in the context of the causal filter estimation, guides tractography through regions with partial volume effects. We validate the method on synthetic data and present results on two types in vivo data: diffusion tensors and a spherical harmonic reconstruction of the fibre orientation distribution function (fODF).

The measurement of the distance between diffusion tensors is the foundation on which any subsequent analysis or processing of these quantities, such as registration, regularization, interpolation, or statistical inference is based. In recent years a family of Riemannian tensor metrics based on geometric considerations has been introduced for this purpose. In this work we examine the properties one would use to select metrics for diffusion tensors, diffusion coefficients, and diffusion weighted MR image data. We show that empirical evidence supports the use of a Euclidean metric for diffusion tensors, based upon Monte Carlo simulations. Our findings suggest that affine invariance is not a desirable property for a diffusion tensor metric because it leads to substantial biases in tensor data. Rather, the relationship between distribution and distance is suggested as a novel criterion for metric selection.

We propose algorithms for tracking the boundary contour of a deforming object from an image sequence, when the nonaffine (local) deformation over consecutive frames is large and there is overlapping clutter, occlusions, low contrast, or outlier imagery. When the object is arbitrarily deforming, each, or at least most, contour points can move independently. Contour deformation then forms an infinite (in practice, very large), dimensional space. Direct application of particle filters (PF) for large dimensional problems is impractically expensive. However, in most real problems, at any given time, most of the contour deformation occurs in a small number of dimensions ("effective basis space") while the residual deformation in the rest of the state space ("residual space") is small. This property enables us to apply the particle filtering with mode tracking (PF-MT) idea that was proposed for such large dimensional problems in recent work. Since most contour deformation is low spatial frequency, we propose to use the space of deformation at a subsampled set of locations as the effective basis space. The resulting algorithm is called deform PF-MT. It requires significant modifications compared to the original PF-MT because the space of contours is a non-Euclidean infinite dimensional space.