ABSTRACT:
Electrocardiogram (ECG) is a non invasive procedure that has been widely used in the diagnosis of irregular rhythm of the heart in the medical domain. The non stationary nature of ECG waveform has made an interesting field of research in biosignal analysis. ECG signal is characterized by QRS complex, P and T waves which forms the basic morphological features. However other features can be extracted from its intervals and amplitudes. The time consuming nature of manual ECG signal analysis, selecting useful and relevant features for disease classification, and high dimensionality of features have been issues addressed in literatures over time. Considering these issues and challenges forms the bane of this paper which centers on a review of ECG feature extraction techniques. Different paradigms and techniques adopted in several studies on ECG feature extraction are compared. This review is intended to give researchers in the field of biosignal analysis insights into the areas well considered and recommend further research directions.
Keywords:
ECG, Feature extraction, Metaheuristics.
