Título: Estratificação de Risco em Pacientes com Infarto Agudo do Miocárdio Através de Redes Neurais Artificiais e Análise de Correspondência
Autores: Zanini, Alexandre; Ladeira, Renata Teixeira; Pedreira, Carlos Eduardo; Nicolau, José Carlos
Resumo: In this article an application of Artificial Neural Networks and Correspondence Analysis, with the goal of forecasting the mortality prognosis in patients with acute myocardial infarction, it is proposed. Our database was originally composed by 150 variables associated to 522 patients with acute myocardial infarction. Correspondence Analysis was applied to find the representative variables that are statistically associated with mortality. Finally, the selected variables were used as inputs for the Kohonen’s algorithm with the goal of finding patients patterns of the patients that evolute to mortality.
Código DOI: 10.21528/CBRN2003-093
Artigo em PDF: 6CBRN_093.PDF
Arquivo BibTex: 6CBRN_093.bib