Neste trabalho apresentamos uma nova abordagem para a exploração de fibras neurais a partir de imagens de tensores de difusão. A estratégia combina técnicas de visualização de informação e visualização científica para obter uma rápida e precisa interpretação das fibras. Para isto fazemos uma transformação das fibras para vetores, e utilizamos uma nova técnica de projeção multi-dimensional (P-LSP), para trabalhar com conjuntos grandes de dados. A exploração do espaço das fibras é feita através desta projeção. Além disso, é apresentada a extensão das técnicas LSP e P-LSP para criar projeções em 3D, assim como estratégias que permitem interagir com pontos em 3D. Outra contribuição deste trabalho é a modificação de um método apresentado para criar superfícies de densidade fechadas sobre pontos esparsos. Esta modificação torna possível criar superfícies sobre conjuntos de pontos maiores com uma qualidade aceitável, o que é utilizado para representar conjunto de fibras como uma superfície. Esta nova abordagem é comparada com trabalhos similares mostrando nossas vantagens em termos de tempo de processamento, qualidade e funcionalidades para analisar esta categoria de dados / This work presents a novel approach for the exploration of neural fibers extracted from Diffusion Tensor Images. The developed strategy combines techniques from information and scientific visualization in order to attain a fast and precise interpretation of fiber sets. The approach transforms fibers into vectors from which a new multidimensional projection technique (PLSP) capable of handling large data sets. The fiber space is explored through the projection. Additionally this work extends P-LSP and LSP projections to 3D, and defines strategies to interact with 3D sparse points. Another contribution of this work is the extension of a method to create close density surfaces over the sparse space generated by the projections. The visualization approach is compared with other similar work showing advantages in processing time, quality and exploration capability to analyze this type of data sets

The majority of current multi-label classification research focuses on learning dependency structures among output labels. This paper provides a novel theoretical view on the purported assumption that effective multi-label classification models must exploit output dependencies. We submit that the flurry of recent dependency-exploiting, multi-label algorithms may stem from the deficiencies in existing datasets, rather than an inherent need to better model dependencies. We introduce a novel categorization of multi-label metrics, namely, evenly and unevenly weighted label metrics. We explore specific features that predispose datasets to improved classification by methods that model label dependence. Additionally, we provide an empirical analysis of 15 benchmark datasets, 1 real-life dataset, and a variety of synthetic datasets. We assert that binary relevance (BR) yields similar, if not better, results than dependency-exploiting models for metrics with evenly weighted label contributions. We qualify this claim with discussions on specific characteristics of datasets and models that render negligible the differences between BR and dependency-learning models.

Numerical modeling of extreme flooding in an urban area in eastern Iowa is presented. Modeling is performed using SRH-2D, an unstructured grid, finite volume model that solves the depth-averaged shallow-water equations. Data from a photogrammetric stereo compilation, contour maps, a hydrographic survey and building records were used to create a digital elevation model depicting the river channel and floodplain. A spatially distributed Manning coefficient based on land cover classification, derived from aerial photography is also used. The model is calibrated with high-resolution inundation depth data derived from a 1 m light detection and ranging survey, collected during the falling limb of the flood hydrograph, and discrete global positioning system measurements of water surface elevation at a bankfull condition. The model is validated with discrete high water marks collected immediately after the flood event. Results show the model adequately represents the water surface elevation in the main channel and floodplain and that exclusion of the discharges from minor creeks did not affect simulation accuracy. Reach scale results are not affected by the presence of buildings, but local inconsistencies occur in shallow water if buildings are not removed from the mesh. An unsteady hydrograph approximates flood hydrodynamics better than a steady-state simulation, but extreme computation time is not feasible for most investigations. The two-dimensional model was also compared to a comparable one-dimensional model of the study reach. The 1D model suffered from an inability to accurately predict inundation depth throughout the entire study area.

Marfan Syndrome leads to the weakening of the thoracic aorta and ultimate rupture causing death of the patient. Current monitoring method involves measuring the diameter of the aorta near the heart. Our approach is to develop a new technology that will provide clinicians the ability to evaluate the size, shape and motion of the entire thoracic aorta using four-dimensional cardiac MRI. This project alters the existing research algorithms to provides an integrated application for processing the images and provides novel measurements about the aorta from a data set of 32 normal subjects and 38 patients with serial scans.

This dissertation introduces a novel three-dimensional pseudo-rigid-body model (3-D PRBM) for straight cantilever beams with rectangular cross sections. The model is capable of capturing the behavior of the neutral axis of a beam loaded with an arbitrary force end-load. Numerical integration of a system of differential equations yields approximate displacement and orientation of the beam's neutral axis at the free end, and curvatures of the neutral axis at the fixed end. This data was used to develop the 3-D PRBM which consists of two torsional springs connecting two rigid links for a total of 2 degrees of freedom (DOF). The 3-D PRBM parameters that are comparable with existing 2-D model parameters are characteristic radius factor (mean: γ = 0.8322), bending stiffness coefficient (mean: KΘ = 2.5167) and parametric angle coefficient (mean: cΘ = 1.2501). New parameters are introduced in the model in order to capture the spatial behavior of the deflected beam, including two parametric angle coefficients (means: cΨ = 1.0714; cΦ = 1.0087). The model is verified in a few locations using ANSYSTM and its use in the design of compliant mechanisms is illustrated through spatial compliant versions of crank slider and double slider mechanisms.

The main purpose of this dissertation was to examine the statistical dependence of imputed microRNA (miRNA) data in a colorectal cancer study. The dissertation addressed three related statistical issues that were raised by this study. the first statistical issue was motivated by the fact that miRNA expression was measured in paired tumor-normal samples of hundreds of patients, but data for many normal samples were missing due to lack of tissue availability. We compared the precision and power performance of several imputation methods, and drew attention to the statistical dependence induced by K-Nearest Neighbors (KNN) imputation. The second statistical issue was raised by the necessity to address the bimodality of distributions of miRNA data along with the imputation-induced dependency among subjects. We proposed and compared the performance of three nonparametric methods to identify the dierentially expressed miRNAs in the paired tumor-normal data while accounting for the imputation-induced dependence. The third statistical issue was related to the development of a normalization method for miRNA data that would reduce not only technical variation but also the variation caused by the characteristics of subjects, while maintaining the true biological dierences between arrays.

Since the first discovery of atomically thin sheets of carbon, two-dimensional (2D) materials have captured the interest from scientific community to expand the understanding in fundamental physics and chemistry at low dimensional systems. With extraordinary phenomena only possible at atomically thin limits, there has been high demand to reveal new and unique 2D materials and manipulate their structures and properties. Structural tunability of superatomic solids motivates us to control dimentionality of the materials and construct layered structures which could be exfoliated to 2D materials. The layered crystal [Co6Se8(PEt2phen)6][C60]5 can be used as a template to create a 2D C60-based material with an optical gap in mid-infrared. Re6Se8Cl2 and Mo6S3Br6, are presented as the first examples of covalently linked 2D superatomic solids built from nanoscale building blocks with hierarchical structures and semiconducting properties. We further demonstrate the emergence of hierarchical coherent phonons in a 2D superatomic semiconductor Re6Se8Cl2. Lastly, we explore complex magnetic phases in 2D ferromagnetic semiconductor CrSBr using second harmonic generation and Raman spectroscopy. 2D superatomic semiconductors and 2D magnetic materials provide additional sets of design principles to manipulate structural, electronic, phononic, and magnetic properties at the atomically thin limits. These materials hold promises as model systems to study fundamental physical principles as well as platform for applications with phonon engineering and magnetic optoelectronic devices.

acase@tulane.edu / Extreme phenotype sampling (EPS) is a broadly-used design to identify candidate genetic factors contributing to the variation of quantitative traits. By enriching the signals in the extreme phenotypic samples within the top and bottom percentiles, EPS can boost the study power compared with the random sampling with the same sample size. The existing statistical methods for EPS data test the variants/regions individually. However, many disorders are caused by multiple genetic factors. Therefore, it is critical to simultaneously model the effects of genetic factors, which may increase the power of current genetic studies and identify novel disease-associated genetic factors in EPS. The challenge of the simultaneous analysis of genetic data is that the number (p ~10,000) of genetic factors is typically greater than the sample size (n ~1,000) in a single study. The standard linear model would be inappropriate for this p>n problem due to the rank deficiency of the design matrix. An alternative solution is to apply a penalized regression method – the least absolute shrinkage and selection operator (LASSO). LASSO can deal with this high-dimensional (p>n) problem by forcing certain regression coefficients to be zero. Although the application of LASSO in genetic studies under random sampling has been widely studied, its statistical inference and testing under EPS remain unknown. We propose a novel sparse model (EPS-LASSO) with hypothesis test for high-dimensional regression under EPS based on a decorrelated score function to investigate the genetic associations, including the gene expression and rare variant analyses. The comprehensive simulation shows EPS-LASSO outperforms existing methods with superior power when the effects are large and stable type I error and FDR control. Together with the real data analysis of genetic study for obesity, our results indicate that EPS-LASSO is an effective method for EPS data analysis, which can account for correlated predictors. / 1 / Chao Xu

A theoretical approach was developed to model the vibrational dynamics of amorphous, two-dimensional materials. The materials were modeled as continuous random networks (CRN's) comprising an assemblage of planar rings of diverse size. In-plane vibrational modes for symmetric 4-, 5-, 60, 7-, and 8-membered rings were examined. Vibrational states of isolated rings were modified by coupling the rings to a continuous network to represent rings embedded in a CRN. An effective force constant was used to couple the ring vibrations to the network's collective motions. Potentials were approximated with the use of a central force model (bond-stretching force constant) and a valence force model (bond-stretching and bond-angle-bending force constants). Valence force model calculations employed group theory. Mode frequencies were calculated using the method of small oscillations and the normal coordinate treatment. Amorphous carbon was used as a test case for the embedded ring approach. A physically consistent set of force constants for the valence force model was determined by comparing the 6-membered ring E2g mode in graphite. Frequencies for selected ring modes were calculated, resulting in a discrete line spectrum. Calculated frequencies were fitted with gaussian peaks and convoluted into theoretical spectra for comparison with the experimental Raman spectrum of amorphous carbon. Integrated gaussian lineshape intensities were assumed to be directly proportional to the CRN ring statistics. The peaks were convoluted with the peak widths, ring statistics, and number of modes as the adjustable parameters. Parameters consistent with previous research on the structure and dynamics of amorphous carbon provided satisfactory fits to the data. The best fit to the Raman data includes the E2g and A1g modes of 6-membered rings (present in Raman spectra of nanocrystalline graphite), and the Raman active E2' modes of 5- and 7-membered rings. The corresponding ring statistics agree with previous results, supporting the presence of a sizable percentage of 5- and 7- membered rings, but with no 4- or 8-membered rings. This positive result provides verification for the embedded ring approach, and supports a CRN model for amorphous carbon.

Text-to-scene generation systems take input in the form of a natural language text and output a 3D scene illustrating the meaning of that text. A major benefit of text-to-scene generation is that it allows users to create custom 3D scenes without requiring them to have a background in 3D graphics or knowledge of specialized software packages. This contributes to making text-to-scene useful in scenarios from creative applications to education. The primary goal of this thesis is to explore how we can use text-to-scene generation in a new way: as a tool to facilitate the elicitation and formal documentation of language. In particular, we use text-to-scene generation (a) to assist field linguists studying endangered languages; (b) to provide a cross-linguistic framework for formally modeling spatial language; and (c) to collect language data using crowdsourcing. As a side effect of these goals, we also explore the problem of multilingual text-to-scene generation, that is, systems for generating 3D scenes from languages other than English. The contributions of this thesis are the following. First, we develop a novel tool suite (the WordsEye Linguistics Tools, or WELT) that uses the WordsEye text-to-scene system to assist field linguists with eliciting and documenting endangered languages. WELT allows linguists to create custom elicitation materials and to document semantics in a formal way. We test WELT with two endangered languages, Nahuatl and Arrernte. Second, we explore the question of how to learn a syntactic parser for WELT. We show that an incremental learning method using a small number of annotated dependency structures can produce reasonably accurate results. We demonstrate that using a parser trained in this way can significantly decrease the time it takes an annotator to label a new sentence with dependency information. Third, we develop a framework that generates 3D scenes from spatial and graphical semantic primitives. We incorporate this system into the WELT tools for creating custom elicitation materials, allowing users to directly manipulate the underlying semantics of a generated scene. Fourth, we introduce a deep semantic representation of spatial relations and use this to create a new resource, SpatialNet, which formally declares the lexical semantics of spatial relations for a language. We demonstrate how SpatialNet can be used to support multilingual text-to-scene generation. Finally, we show how WordsEye and the semantic resources it provides can be used to facilitate elicitation of language using crowdsourcing.