Analysis of Changes in the Beat-to-Beat P-wave Morphology Using Clustering Techniques

In English Several pathologies related to the atrial electrical activity can be detected in the electrocardiogram P-wave. A protocol for analyzing P-wave morphology changes has been developed in this article. By using this protocol a study on the beat-to-beat P-wave morphology changes of 89 ECG signals is performed. An
algorithmbased on the embedding space techniques has been used to extract the P-wave information of the ECG. The P-waves obtained in several of these ECGs exhibit significant alternate morphology changes. The morphologies have been classified by using the K-means clustering algorithm. The mechanism behind the P-wave morphology change process and its possible pathophysiological importance remains to be clarified.