Descripción
|
|
---|---|
Chronic diseases require ongoing care to improve patients? quality of life. Large amounts of public and private investment are consumed in dealing with issues like employee absenteeism, early retirement and social spending. Nowadays, it is estimated that 12% of natural deaths occur suddenly of which 88% are of cardiac origin. Early heart beat anomalies detection plays a key role in preventing cardiac diseases. This paper proposes the use of time series data mining to extract relevant electrocardiogram (ECG) features to predict the probability of ventricular fibrillation (VF) events. Decision trees, k-nearest neighbors, support vector machines, logistic regression and neural networks have been applied to ECG data. Different feature sets have been proposed and evaluated combining different beat sequences lengths (1, 3, 6 or 9 beats), ECG data points (P, Q, R, S, T) and segments (PS, QT, ST, PR and RR). These data mining models could be implemented in computer-aided diagnosis (CAD) systems to evaluate long-term ECG data of a patient and identify VF events in advance. | |
Internacional
|
Si |
Nombre congreso
|
2019 IEEE 32nd International Symposium on Computer-Based Medical Systems (CBMS) |
Tipo de participación
|
960 |
Lugar del congreso
|
Cordoba, Spain |
Revisores
|
Si |
ISBN o ISSN
|
978-1-7281-2286-1 |
DOI
|
10.1109/CBMS.2019.00014 |
Fecha inicio congreso
|
05/06/2019 |
Fecha fin congreso
|
07/06/2019 |
Desde la página
|
14 |
Hasta la página
|
19 |
Título de las actas
|
al ISBN o ISSN: 2019 IEEE 32nd International Symposium on Computer-Based Medical Systems (CBMS) |