ECG signal feature extraction and classification based on R peaks detection in the phase space

The goal of this paper is to present a novel approach in the automatic diagnosis of ECG abnormalities based on detection of R peaks in the phase space. The features are extracted from detected R peaks using their geometric position on the phase curve. This paper is dealing with classification problem of normal and abnormal ECG signals. The proposed system has been validated with the data from the MIT-BIH database, in order to detect the cardiac arrhythmia. Support Vector Machine and K-Nearest Neighbour are used as classifiers. Results for both classifiers are similar. They are showing high accuracy in the experiment of classifying one test signal.