ECG signal processing using classifier to analyses cardiovascular disease

摘要

Commonly used methods in biomedical engineering research is Digital signal processing and data analysis. Employment of digital signal filtering on electrocardiogram (ECG) and classification is according to their features that utilizes support vector machine (SVM) in this paper. Automatic detection and classification of noises can improvise the noise containing electrocardiogram (ECG). This concept consists of three stages, namely ECG signal pre-processing, feature selection, and classification of ECG beats. To remove noise from ECG, filters are developed. Designed filters are focused on removing artefacts. Moreover, this paper contains SVM algorithm that consists of an automatic classifier to detect the five pathologies. The classifier was tested on the physioNet ECG Database with an accuracy of 96.60 %.