Thesis (Ph. D.)--Ohio State University, 2005. / Title from first page of PDF file. Includes bibliographical references (p. 129-145).

Made available in DSpace on 2014-06-11T19:27:11Z (GMT). No. of bitstreams: 0 Previous issue date: 2006-02-24Bitstream added on 2014-06-13T19:55:27Z : No. of bitstreams: 1 costa_raa_me_guara.pdf: 2052235 bytes, checksum: a7679d1faabaecae7d1f08931905e615 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / A pesquisa desenvolvida para reconhecimento de padrões estatísticos com base nas informações de textura, ainda possui muitos desafios a serem superados. Um deles é a extração de características a partir de imagens que apresentam pouca variação de contraste. Este trabalho propõe um algoritmo de seleção de características de textura baseado nas característica de Haralick como a correlação e a entropia. A proposta dessa abordagem é tornar possível classificar os dados, linearmente separáveis ou não, em duas ou mais classes levando em conta um pequeno subespaço de características. Foram estudadas técnicas estatísticas para obtenção da textura, para análise e quantificação de colônias A e a fração volumétrica de Widmanstatten, presentes na liga Ti-6Al-4v. A pesquisa culminou no desenvolvimento de um algoritmo e na realização de testes de caracterização baseadas na textura das imagens. No primeiro teste utilizou-se a correlação entre duas áreas, de dimensões diferentes, de uma mesma imagem para identificação de pontos homólogos entre essas imagens onde foi possível a obtenção da fração volumétrica das fases A e B, onde a variação tonal é evidente. Para o segundo teste utilizou-se a correlação entre os mapas de entropia, seguindo o mesmo procedimento anterior para obter os pontos homólogos entre as imagens. A metodologia proposta pode ser aplicada a outros tipos de materiais. A partir dos resultados da fração de área A e a fração de área das colônias A, poderão ser feito ensaios nos quais será possível relacionar as porcentagens citadas com as propriedades mecânicas da liga. / The research developed for recognition of statistical patterns based on texture information has still many challenges to be overcome. One of them is characteristics extraction from images which display little contrast variation. This work proposes and algorithm of texture characteristics selection based on Haralick characteristics as correlation and entropy. This iniciation proposal is to make possible to classify data, linearly separable or not, in two or more ranges taking in account a little area of characteristics. It has been studied statistical techniques to obtain texture, for analysis and amount of A colony and Widmanstatten volumeter fraction, present in Ti-6Al-4V alloy. This research has culminated in an algorithm development and in characterization tests accomplishment based on image texture. In the first test there has been used correlation between two areas of different dimension, from the same picture for similar points identification between these images where it has been possible to obtain volumetric fraction of phases A and B where to tonal variation is evident. For the second test it has used correlation between entropy maps, following the previous procedure in order to obtain similar points between the images. The proposed methodology can be applied to other material types. From results of area fraction and A colonies area fraction it can be made tests in which it will be possible to relate the percentages cited with the alloy mechanical properties.

Orientador: Luis Rogério de Oliveira Hein / Banca: Fernando de Azevedo Silva / Banca: Carlos Frederico de Angelis / Resumo: A pesquisa desenvolvida para reconhecimento de padrões estatísticos com base nas informações de textura, ainda possui muitos desafios a serem superados. Um deles é a extração de características a partir de imagens que apresentam pouca variação de contraste. Este trabalho propõe um algoritmo de seleção de características de textura baseado nas característica de Haralick como a correlação e a entropia. A proposta dessa abordagem é tornar possível classificar os dados, linearmente separáveis ou não, em duas ou mais classes levando em conta um pequeno subespaço de características. Foram estudadas técnicas estatísticas para obtenção da textura, para análise e quantificação de colônias A e a fração volumétrica de Widmanstatten, presentes na liga Ti-6Al-4v. A pesquisa culminou no desenvolvimento de um algoritmo e na realização de testes de caracterização baseadas na textura das imagens. No primeiro teste utilizou-se a correlação entre duas áreas, de dimensões diferentes, de uma mesma imagem para identificação de pontos homólogos entre essas imagens onde foi possível a obtenção da fração volumétrica das fases A e B, onde a variação tonal é evidente. Para o segundo teste utilizou-se a correlação entre os mapas de entropia, seguindo o mesmo procedimento anterior para obter os pontos homólogos entre as imagens. A metodologia proposta pode ser aplicada a outros tipos de materiais. A partir dos resultados da fração de área A e a fração de área das colônias A, poderão ser feito ensaios nos quais será possível relacionar as porcentagens citadas com as propriedades mecânicas da liga. / Abstract: The research developed for recognition of statistical patterns based on texture information has still many challenges to be overcome. One of them is characteristics extraction from images which display little contrast variation. This work proposes and algorithm of texture characteristics selection based on Haralick characteristics as correlation and entropy. This iniciation proposal is to make possible to classify data, linearly separable or not, in two or more ranges taking in account a little area of characteristics. It has been studied statistical techniques to obtain texture, for analysis and amount of A colony and Widmanstatten volumeter fraction, present in Ti-6Al-4V alloy. This research has culminated in an algorithm development and in characterization tests accomplishment based on image texture. In the first test there has been used correlation between two areas of different dimension, from the same picture for similar points identification between these images where it has been possible to obtain volumetric fraction of phases A and B where to tonal variation is evident. For the second test it has used correlation between entropy maps, following the previous procedure in order to obtain similar points between the images. The proposed methodology can be applied to other material types. From results of area fraction and A colonies area fraction it can be made tests in which it will be possible to relate the percentages cited with the alloy mechanical properties. / Mestre

[pt] O presente trabalho visa desenvolver uma ferramenta computacional para sugerir sobre a malignidade ou benignidade de Nódulos Pulmonares Solitários, através da análise de medidas de textura e geometria obtidas a partir das imagens de tomografia computadorizada. São propostos quatro grupos de métodos com o objetivo de sugerir o diagnóstico para o nódulo. Os grupos de métodos são divididos de acordo com suas características comuns. O Grupo I trata dos métodos baseados em textura adaptados para 3D, como o histograma, o Método de Dependência Espacial de Níveis de Cinza, o Método de Diferença de Níveis de Cinza e o Método de Comprimento de Primitivas de Níveis de Cinza. O Grupo II também trata da textura dos nódulos, mas utiliza quatro funções geoestatísticas denominadas semivariograma, semimadograma, covariograma e correlograma. O Grupo III descreve apenas medidas baseadas na geometria do nódulo, como a convexidade, a esfericidade e medidas baseadas na curvatura. Por fim, o Grupo IV analisa os métodos do coeficiente de Gini e do esqueleto dos nódulos, que levam em consideração tanto a geometria quanto a textura do nódulo. Foi analisada uma amostra com 36 nódulos, sendo 29 benignos e 7 malignos, e os resultados preliminares são promissores na caracterização dos nódulos pulmonares. A maioria dos grupos de métodos propostos tem o valor da área sobre a curva ROC acima de 0.800, utilizando a Análise Discriminante Linear de Fisher e a Rede Neural Perceptron de Múltiplas Camadas. Isto significa que os métodos propostos possuem grande potencial na discriminação e classificação dos Nódulos Pulmonares Solitários. / [en] The present work seeks to develop a computational tool to suggest about the malignancy or benignity of Solitary Lung Nodules by the analysis of texture and geometry measures obtained from computadorized tomography images. Four groups of methods are proposed with the purpose of suggesting the diagnosis for such nodule. The groups of methods are divided according to their common characteristics. Group I includes methods based on texture adapted for 3D, such as the histogram, the Spatial Gray Level Dependence Method, the Gray Level Difference Method and Gray Level Run Length Matrices. Group II also deals with the texture of nodules, but uses four statistical functions denominated semivariogram, semimadogram, covariogram and correlogram. Group III describes measures based only on the geometry of the nodule, such as convexity, sphericity, and measures based on the curvature. Finally, Group IV analyzes the Gini coeficient and nodule skeleton methods, which take into account both the nodule s geometry and its texture. A sample with 36 nodules, 29 benign and 7 malignant, was analyzed and the preliminary results of this approach are very promising in characterizing lung nodules. Most groups of proposed methods have the area under the ROC curve value above 0.800, using Fisher s Linear Discriminant Analysis and Multilayer Perceptron Neural Networks. This means that the proposed methods have great potential in the discrimination and classification of Solitary Lung Nodules.

Cette thèse porte sur des canevas méthodologiques paramétriques pour la modélisation de champs de tenseurs de structure locaux (TSL) calculés sur des images texturées. Estimé en chaque pixel, le tenseur de structure permet la caractérisation de la géométrie d’une image texturée à travers des mesures d’orientation et d’anisotropie locales. Matrices symétriques semi-définies positives, les tenseurs de structure ne peuvent pas être manipulés avec les outils classiques de la géométrie euclidienne. Deux canevas statistiques riemanniens, reposant respectivement sur les espaces métriques a ne invariant (AI) et log-euclidien (LE), sont étudiés pour leur représentation. Dans chaque cas, un modèle de distribution gaussienne et de mélange associé sont considérés pour une analyse statistique. Des algorithmes d’estimation de leurs paramètres sont proposés ainsi qu’une mesure de dissimilarité. Les modèles statistiques proposés sont tout d’abord considérés pour décrire des champs de TSL calculés sur des images texturées. Les modèles AI et LE sont utilisés pour décrire des distributions marginales de TSL tandis que les modèles LE sont étendus afin de décrire des distributions jointes de TSL et de caractériser des dépendances spatiales et multi-échelles. L’ajustement des modèles théoriques aux distributions empiriques de TSL est évalué de manière expérimentale sur un ensemble de textures composées d’un spectre assez large de motifs structuraux. Les capacités descriptives des modèles statistiques proposés sont ensuite éprouvées à travers deux applications. Une première application concerne la reconnaissance de texture sur des images de télédétection très haute résolution et sur des images de matériaux carbonés issues de la microscopie électronique à transmission haute résolution. Dans la plupart des cas, les performances des approches proposées sont supérieures à celles obtenues par les méthodes de l’état de l’art. Sur l’espace LE, les modèles joints pour la caractérisation des dépendances spatiales au sein d’un champ de TSL améliorent légèrement les résultats des modèles opérant uniquement sur les distributions marginales. La capacité intrinsèque des méthodes basées sur le tenseur de structure à prendre en considération l’invariance à la rotation, requise dans beaucoup d’applications portant sur des textures anisotropes, est également démontrée de manière expérimentale. Une deuxième application concerne la synthèse de champs de TSL. A cet e et, des approches mono-échelle ainsi que des approches pyramidales multi-échelles respectant une hypothèse markovienne sont proposées. Les expériences sont effectuées à la fois sur des champs de TSL simulés et sur des champs de TSL calculés sur des textures réelles. Efficientes dans quelques configurations et démontrant d’un potentiel réel de description des modèles proposés, les expériences menées montrent également une grande sensibilité aux choix des paramètres qui peut s’expliquer par des instabilités d’estimation sur des espaces de grande dimension. / This thesis proposes and evaluates parametric frameworks for modelling local structure tensor (LST) fields computed on textured images. A texture’s underlying geometry is described in terms of orientation and anisotropy, estimated in each pixel by the LST. Defined as symmetric non-negative definite matrices, LSTs cannot be handled using the classical tools of Euclidean geometry. In this work, two complete Riemannian statistical frameworks are investigated to address the representation of symmetric positive definite matrices. They rely on the a ne-invariant (AI) and log-Euclidean (LE) metric spaces. For each framework, a Gaussian distribution and its corresponding mixture models are considered for statistical modelling. Solutions for parameter estimation are provided and parametric dissimilarity measures between statistical models are proposed as well. The proposed statistical frameworks are first considered for characterising LST fields computed on textured images. Both AI and LE models are first employed to handle marginal LST distributions. Then, LE models are extended to describe joint LST distributions with the purpose of characterising both spatial and multiscale dependencies. The theoretical models’ fit to empirical LST distributions is experimentally assessed for a texture set composed of a large diversity of patterns. The descriptive potential of the proposed statistical models are then assessed in two applications. A first application consists of texture recognition. It deals with very high resolution remote sensing images and carbonaceous material images issued from high resolution transmission electron microscopy technology. The LST statistical modelling based approaches for texture characterisation outperform, in most cases, the state of the art methods. Competitive texture classification performances are obtained when modelling marginal LST distributions on both AI and LE metric spaces. When modelling joint LST distributions, a slight gain in performance is obtained with respect to the case when marginal distributions are modelled. In addition, the LST based methods’ intrinsic ability to address the rotation invariance prerequisite that arises in many classification tasks dealing with anisotropic textures is experimentally validated as well. In contrast, state of the art methods achieve a rather pseudo rotation invariance. A second application concerns LST field synthesis. To this purpose, monoscale and multiscale pyramidal approaches relying on a Markovian hypothesis are developed. Experiments are carried out on toy LST field examples and on real texture LST fields. The successful synthesis results obtained when optimal parameter configurations are employed, are a proof of the real descriptive potential of the proposed statistical models. However, the experiments have also shown a high sensitivity to the parameters’ choice, that may be due to statistical inference limitations in high dimensional spaces.

No description available.

Made available in DSpace on 2014-10-09T12:36:19Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T13:59:16Z (GMT). No. of bitstreams: 1 12908.pdf: 2121499 bytes, checksum: c2d2abf4d18418855c9e93a43337b8ad (MD5) / Dissertacao (Mestrado) / IEA/D / Escola Politecnica, Universidade de Sao Paulo - POLI/USP

Made available in DSpace on 2014-10-09T12:50:39Z (GMT). No. of bitstreams: 0 / Made available in DSpace on 2014-10-09T13:58:59Z (GMT). No. of bitstreams: 1 00051.pdf: 1323615 bytes, checksum: 560d14914e0156e890a27864c7db81e2 (MD5) / Dissertacao (Mestrado) / IEA/D / Instituto de Fisica, Universidade de Sao Paulo - IF/USP

abstract: Hepatocellular carcinoma (HCC) is a malignant tumor and seventh most common cancer in human. Every year there is a significant rise in the number of patients suffering from HCC. Most clinical research has focused on HCC early detection so that there are high chances of patient's survival. Emerging advancements in functional and structural imaging techniques have provided the ability to detect microscopic changes in tumor micro environment and micro structure. The prime focus of this thesis is to validate the applicability of advanced imaging modality, Magnetic Resonance Elastography (MRE), for HCC diagnosis. The research was carried out on three HCC patient's data and three sets of experiments were conducted. The main focus was on quantitative aspect of MRE in conjunction with Texture Analysis, an advanced imaging processing pipeline and multi-variate analysis machine learning method for accurate HCC diagnosis. We analyzed the techniques to handle unbalanced data and evaluate the efficacy of sampling techniques. Along with this we studied different machine learning algorithms and developed models using them. Performance metrics such as Prediction Accuracy, Sensitivity and Specificity have been used for evaluation for the final developed model. We were able to identify the significant features in the dataset and also the selected classifier was robust in predicting the response class variable with high accuracy. / Dissertation/Thesis / M.S. Industrial Engineering 2013

No description available.