Authors:
Diana Batista
1
and
Ana Fred
2
Affiliations:
1
Instituto Superior Técnico, Portugal
;
2
Instituto Superior Técnico and Instituto de Telecomunicações / IST, Portugal
Keyword(s):
Atrial fibrillation, ECG, Wavelet, Pattern Analysis, Artificial Neural Network, k-Nearest Neighbours, Support Vector Machine.
Related
Ontology
Subjects/Areas/Topics:
Applications and Services
;
Artificial Intelligence
;
Biomedical Engineering
;
Biomedical Signal Processing
;
Computer Vision, Visualization and Computer Graphics
;
Data Manipulation
;
Health Engineering and Technology Applications
;
Human-Computer Interaction
;
Medical Image Detection, Acquisition, Analysis and Processing
;
Methodologies and Methods
;
Neurocomputing
;
Neurotechnology, Electronics and Informatics
;
Pattern Recognition
;
Physiological Computing Systems
;
Sensor Networks
;
Soft Computing
;
Wavelet Transform
Abstract:
Atrial fibrillation (AF) is the most common type of arrhythmia. This work presents a pattern analysis
approach to automatically classify electrocardiographic (ECG) records as normal sinus rhythm or AF. Both
spectral and time domain features were extracted and their discrimination capability was assessed
individually and in combination. Spectral features were based on the wavelet decomposition of the signal
and time parameters translated heart rate characteristics. The performance of three classifiers was evaluated:
k-nearest neighbour (kNN), artificial neural network (ANN) and support vector machine (SVM). The MITBIH
arrhythmia database was used for validation. The best results were obtained when a combination of
spectral and time domain features was used. An overall accuracy of 99.08 % was achieved with the SVM
classifier.