Estimativa da temperatura do cérebro isquêmico através do método do filtro de partículas

• Ano de defesa: 2019
• Autor(a) principal: Nunes, Felipe Sant’Anna
• Tipo de documento: Dissertação
• 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 Mecânica; UFRJ
• Palavras-chave em Português: Encefalopatia Hipóxico-Isquêmica; Problema inverso; Filtro de partículas; CNPQ::ENGENHARIAS::ENGENHARIA MECANICA
• Link de acesso: http://hdl.handle.net/11422/13527

In this work, a two-dimensional model based on the Pennes equation was developed to analyze the behavior of the temperature distribution in the head of a newborn under cooling induced by the use of a cap. Three cases were considered, including: healthy brain condition, local ischemic condition and total ischemic condition. For the ischemic regions, the blood perfusion was reduced by 80% of its normal value. In the other regions, the perfusion was considered as temperature dependent. The numerical solution for the temperature field was obtained using the finite volume method. The computational code was implemented in Matlab and its verification was made by comparing the results obtained for the direct problem with the results obtained in an equivalent bioheat transfer model built in COMSOL. The verification of the solution of the direct problem was made through the analysis of the mesh convergence and through an energy balance. The solution of the state estimation problem was obtained with the sampling importance resampling algorithm, SIR, of the particle filter method. Noninvasive synthetic temperature measurements, at three points on the surface of the head and at a point on the exit of the cooling cap, were considered available. It was noticed that, depending on the uncertainties associated with the state evolution model and with the observation model, the algorithm was able to estimate the expected exact temperatures. As for the regions where the measurements were available, the algorithm exhibited good accuracy in the estimates for all cases studied