{"id":145,"date":"2017-06-23T11:31:23","date_gmt":"2017-06-23T14:31:23","guid":{"rendered":"http:\/\/web.inf.ufpr.br\/vri2\/?page_id=145"},"modified":"2017-06-23T11:31:23","modified_gmt":"2017-06-23T14:31:23","slug":"luiza-dri-bagesteiro-msc-2015","status":"publish","type":"page","link":"https:\/\/web.inf.ufpr.br\/vri\/alumni\/luiza-dri-bagesteiro-msc-2015\/","title":{"rendered":"Luiza Dri Bagesteiro, Msc, 2015"},"content":{"rendered":"

Classifica\u00e7\u00e3o de Padr\u00f5es Radiol\u00f3gicos Por Blocos em Imagens N\u00e3o Segmentadas de Tomografia Computadorizada – Blockwise Classification of Radiological Patterns in Non-Segmented Images of Computed Tomography<\/h2>\n
<\/div>\n
Author<\/strong>:\u00a0Luiza Dri Bagesteiro (Lattes<\/a><\/span>)<\/strong><\/div>\n
\n
\n

E-mail: ldbagesteiro@inf.ufpr.br, lu.bagesteiro@gmail.com<\/strong><\/p>\n

Supervisor:\u00a0<\/strong>Daniel Weingaertner<\/a><\/span><\/strong><\/p>\n

Co-supervisor:<\/strong>\u00a0Lucas Ferrari de Oliveira<\/a><\/span><\/strong><\/p>\n

Abstract:<\/strong><\/p>\n

Diffuse lung diseases (DLDs) are comprised of more than 180 diseases that cause damage to the lungs. Patients who have these diseases present changes in lung tissue, which can be seen on Computed Tomography (CT) images. Thus, it becomes important to identify alterations in these images because, along with other information, they can define the type of disorder that the patient has. The aim of this study is to automatically recognize radiological patterns in non-segmented lung CT images, and also to classify the areas outside the lung region, in addition to five lung patterns: normal, emphysema, ground-glass opacity, fibrosis and micronodules. To achieve the objectives, a methodology that evaluates various descriptors is proposed. Searching for a complementarity of the results, combinations are performed among the classifiers. Using a public dataset containing 20.540 CT blocks, each descriptor was subjected to a classifier Support Vector Machine generating scores to represent the likelihood of the samples belonging to each of the six classes. After that, the classifiers were combined using fusion rules. The rotation invariant uniform CLBP classifier achieved the best individual result, and combined with top-hat transform descriptor, reached rates of 81.65% of sensitivity and 96.06% of specificity. Based on the relevance of the results obtained, the next step is to analyze the possibility of developing a support tool for diagnosis using the developed methodology.<\/p>\n

Resumo:<\/strong><\/p>\n

As Doen\u00e7as Pulmonares Difusas (DPDs) s\u00e3o compostas por mais de 180 patologias que causam danos nos pulm\u00f5es. Os pacientes que possuem essas doen\u00e7as apresentam altera\u00e7\u00f5es no tecido pulmonar, que podem ser visualizadas em imagens de Tomografia Computadorizada (TC). Dessa forma, torna-se importante identificar essas altera\u00e7\u00f5es nas imagens, pois juntamente com outras informa\u00e7\u00f5es, elas podem definir o tipo de doen\u00e7a que o paciente possui. Com base nisso, o objetivo deste trabalho \u00e9 reconhecer automaticamente padr\u00f5es radiol\u00f3gicos em imagens n\u00e3o segmentadas de TC do pulm\u00e3o, e tamb\u00e9m classificar \u00e1reas externas \u00e0 regi\u00e3o pulmonar, al\u00e9m dos cinco padr\u00f5es pulmonares: normal, enfisema, opacidade em vidro-fosco, fibrose e micron\u00f3dulos. Para atingir os objetivos, uma metodologia que avalia diversos descritores \u00e9 proposta. Em busca de uma complementaridade entre os resultados, s\u00e3o realizadas combina\u00e7\u00f5es entre as respostas dos classificadores. Utilizando uma base de imagens p\u00fablica, com 20.540 blocos de imagens de TC, cada descritor foi submetido a um classificador\u00a0Support Vector Machine<\/i>\u00a0que gerou\u00a0scores<\/i>\u00a0para representar a probabilidade das amostras pertencerem a cada uma das seis classes. Ap\u00f3s isso, os classificadores foram combinados empregando algumas regras de fus\u00e3o. O classificador CLBP invariante \u00e0 rota\u00e7\u00e3o e uniforme obteve o melhor resultado individual, e combinado com a transforma\u00e7\u00e3o top-hat, atingiram taxas de 81,65% de sensibilidade e 96,06% de especificidade. Com base na relev\u00e2ncia dos resultados obtidos, o pr\u00f3ximo passo \u00e9 analisar a possibilidade de desenvolver uma ferramenta de aux\u00edlio ao diagn\u00f3stico utilizando a metodologia desenvolvida.<\/p>\n

Related Topics:<\/strong><\/h3>\n