Reconhecimento de padrão em pacientes com esclerose sistêmica por sistemas

• Ano de defesa: 2017
• Autor(a) principal: Oliveira, Fernando Moraes de
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Rio de Janeiro; Brasil; Instituto Alberto Luiz Coimbra de Pós-Graduação e Pesquisa de Engenharia; Programa de Pós-Graduação em Engenharia de Sistemas e Computação; UFRJ
• Palavras-chave em Português: Engenharia de Sistemas e Computação; Processamento de imagem; Sistemas fuzzy; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://hdl.handle.net/11422/10160

Medical images are used to gather information for diagnosis and/or follow-up of patients state. Image acquisition technologies have evolved along the years and the most common are: computed tomography (CT), ultrasonography, magnetic resonance imaging (MRI), nuclear medicine and x-ray. Chest X-ray images invariably have intensity gradations and uncertainties, which leads to the choice of the fuzzy systems (FS) approach. The objective of this study was to optimize images of simple chest X-ray that are used in pulmonary involvement diagnoses' confirmation and follow-up in patients with systemic sclerosis (SS). Pattern recognition using fuzzy methodology, with emphasis on the images segmentation by the intuitionist fuzzy sets (IFS), led to the creation of the pulmonary fibrosis intensity index (P F II) using fuzzy sets (FS). This index, associated with results from pulmoray function tests (PFT), forced vital capacity (F V C) and carbon monoxide difusion capacity (DLCO), assisted on the clinical follow-up of patients with SS. The techniques and methods implemented allowed the development of the SisRPIP - Pattern Recognition System in Pulmonary Imaging. The viability of the SisRPIP was verified with 40 patients' plain chest radiographs already diagnosed with SS, and the results and methodology used, are presented in this thesis.