Premature Ventricular Contraction Beat Detection based on Symbolic Dynamics Analysis

L. Zhao, M. Wiggins, and G. Vachtsevanos (USA)

Keywords

ECG Premature Ventricular ContractonSymbolic Dynamics

Abstract

Premature Ventricular Contraction (PVC) beats are of great importantance in evaluating and predicting life threatening ventricular arrhythmias. In this paper, an algorithm is proposed for the identificaton of PVC beats. ECG signal is first conveted into a symbolic string sequence of `0's and `1's using a thresholding algorithm; Secondly, the length of all substrings composed only with the symbol '1' is estimated; thirdly, the lengths of these substrings are theresholded to detect the PVC beats. Compared to other PVC beat detectors, the proposed algorithm decreases the computational burden, and furthermore, doesn't require for QRS complex detection or calculation of morphological features of the ECG signal. The proposed algorithm has a gross sensitivity of 98.8% for ten patients from the MIT Arrhythmia Database.

Important Links:



Go Back