Classificador Automático e Não-Supervisionado de Batimentos Cardíacos Baseado no Algoritmo Dynamic Tiime Warping

• Ano de defesa: 2008
• Autor(a) principal: Aguiar, Rogério Oliveira de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Espírito Santo; BR; Mestrado em Engenharia Elétrica; Centro Tecnológico; UFES; Programa de Pós-Graduação em Engenharia Elétrica
• Palavras-chave em Português: Eletrocardiografia; Processamento de sinais; Batimento cardíaco; Medidas Elétricas, Magnéticas e Eletrônicas; Instrumentação; 621.3
• Link de acesso: http://repositorio.ufes.br/handle/10/4066

The Telecardio Project is a research project on cardiac patient telemonitoring and automatic detection of risk situations. In this context, an electrocardiogram (ECG) analysis system has been developed as a tool to help medical diagnosis. The system classifies the beats of an ambulatory ECG record using as a reference the patient dominant heartbeat. This classification is due to an original non supervised approach which uses the Dynamic Time Warping method to compare the difference of morphology between heartbeats. Furthermore, the premature heartbeat classification problem is analyzed in this work through the study of the beat annotations made by cardiologist in the MIT-Arrhythmia Database. Finally, a graphic user interface has been proposed to show the results obtained from the ECG analysis system, where useful information and predominant heartbeat morphology along the ECG are exposed to the specialist. Predominant beats are determined by an original algorithm which makes the computation of an average heartbeat from a group of beats belonging to the same class. The system was tested in the MIT-BIH Arrhythmia Database and the results have confirmed the good efficiency of the proposed solution.