Deteccão de complexos QRS em eletrocardiogramas baseada na decomposição em valores singulares em multirresolução

Detalhes bibliográficos
Ano de defesa: 2015
Autor(a) principal: Oliveira, Bruno Rodrigues de [UNESP]
Orientador(a): Não Informado pela instituição
Banca de defesa: Não Informado pela instituição
Tipo de documento: Dissertação
Tipo de acesso: Acesso aberto
Idioma: por
Instituição de defesa: Universidade Estadual Paulista (Unesp)
Programa de Pós-Graduação: Não Informado pela instituição
Departamento: Não Informado pela instituição
País: Não Informado pela instituição
Palavras-chave em Português:
Link de acesso: http://hdl.handle.net/11449/134101
http://www.athena.biblioteca.unesp.br/exlibris/bd/cathedra/14-01-2016/000857500.pdf
Resumo: The electrocardiogram (ECG) is a test of utmost importance to evaluate the heart health of a person. The analysis of its layout enables diagnose a number of diseases and abnor- malities. For computer systems that need to segment the ECG signal, the first step the detection of QRS complexes. The earliest methods used filtering techniques and differen- tiation. Recently some researchers have designed methodologies based the multiresolution analysis the wavelet transform. Following similar path, this work proposes a methodology based on singular value decomposition in multiresolution (MRSVD), to solve the prob- lem of detection of QRS complexes of an ECG signal, which can be divided in four key stages: preprocessing, inspection, decision and correction. In of all, there are innovative proposals: in the first, it is used, in an unprecedented manner, the MRSVD; the second stage explores the physiological characteristics of the heart muscle, to make sure that a particular segment comprises or not a genuine R wave and, the last, corrects detections based on the amplitudes of the R wave. The evaluation of quality of detection for the experiments shows that the proposed algorithm has reached a predictive and sensitivity rates of of 99.20% and 99.70%, respectively, and error of only 1.10%, in comparation to another method which obtained 97, 10%, 99, 00% and 3, 70%, respectively