| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 2.6 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
A Insuficiência Cardíaca Congestiva (ICC) é uma condição crônica em que o coração não bombeia sangue eficientemente. Esta patologia causa fadiga, dispneia, edemas, náuseas e problemas de memória, afetando a qualidade de vida dos pacientes. As causas incluem doenças das artérias coronárias, cardiomiopatia, hipertensão arterial e miocardites. O diagnóstico é realizado com base no histórico do paciente, exames físicos, ecocardiograma, eletrocardiograma e outros métodos. Com a finalidade de aprimorar as ferramentas de diagnóstico, o presente trabalho propõe um modelo de inteligência artificial baseado em deep learning para classificar sinais de ECG patológicos positivos para ICC. Os modelos selecionados foram LSTM e CNN. O treinamento foi realizado com um dataset personalizado criado a partir das bases de dados públicas BIDMC Congestive Heart Failure e PTB Diagnostic ECG da Physionet. Foram selecionados dados de ECG de 28 pessoas, entre 22 e 71 anos, das quais 14 possuem ICC severa (classe NYHA 3 e 4) e 14 amostras de
controle, que não apresentam anormalidades no ECG. A arquitetura da base de dados foi projetada para que o input das redes neurais fossem sinais de ECG puros, sem qualquer tipo de filtragem ou extração de características. Os resultados obtidos demonstraram uma precisão de 98,21% para o modelo CNN e 92,26% para o modelo LSTM.
Congestive Heart Failure (CHF) is a chronic condition in which the heart does not pump blood efficiently. This pathology causes fatigue, dyspnea, edema, nausea, and memory problems, affecting patients’ quality of life. The causes include coronary artery disease, cardiomyopathy, arterial hypertension, and myocarditis. The diagnosis is based on the patient’s medical history, physical exams, echocardiogram, electrocardiogram, and other methods. Aiming to improve diagnostic tools, this study proposes an artificial intelligence model based on deep learning to classify pathological ECG signals indicative of CHF. The selected models were LSTM and CNN. The training was conducted with a personalized dataset created from the public databases BIDMC Congestive Heart Failure and PTB Diagnostic ECG from Physionet. ECG data from 28 individuals, aged 22 to 71, were selected, including 14 with severe CHF (NYHA class 3 and 4) and 14 control samples without ECG abnormalities. The database architecture was designed so that the input to the neural networks was raw ECG signals, without any filtering or feature extraction. The results showed an accuracy of 98.21% for the CNN model and 92.26% for the LSTM model.
Congestive Heart Failure (CHF) is a chronic condition in which the heart does not pump blood efficiently. This pathology causes fatigue, dyspnea, edema, nausea, and memory problems, affecting patients’ quality of life. The causes include coronary artery disease, cardiomyopathy, arterial hypertension, and myocarditis. The diagnosis is based on the patient’s medical history, physical exams, echocardiogram, electrocardiogram, and other methods. Aiming to improve diagnostic tools, this study proposes an artificial intelligence model based on deep learning to classify pathological ECG signals indicative of CHF. The selected models were LSTM and CNN. The training was conducted with a personalized dataset created from the public databases BIDMC Congestive Heart Failure and PTB Diagnostic ECG from Physionet. ECG data from 28 individuals, aged 22 to 71, were selected, including 14 with severe CHF (NYHA class 3 and 4) and 14 control samples without ECG abnormalities. The database architecture was designed so that the input to the neural networks was raw ECG signals, without any filtering or feature extraction. The results showed an accuracy of 98.21% for the CNN model and 92.26% for the LSTM model.
Description
Mestrado de dupla diplomação com a UTFPR - Universidade Tecnológica Federal do Paraná
Keywords
Inteligência artificial Insuficiência cardíaca congestiva Deep learning Eletrocardiograma (ECG)