Analysis of ECG records using ECG Chaos Extractor platform and Weka system

Clustering and classification of ECG records for four patient classes from the Internet databases by using the Weka system. Patient classes include normal, atrial arrhythmia, supraventricular arrhythmia and CHF. Chaos features are extracted automatically by using the ECG Chaos Extractor platform and recorded in Arff files. The list of features includes: correlation dimension, central tendency measure, spatial filling index and approximate entropy. Both ECG signal files and ECG annotations files are analyzed. The results show that chaos features can successfully cluster and classify the ECG annotations records by using standard and efficient algorithms such as EM and C4.5.

[1]  Patrick E. McSharry,et al.  Advanced Methods And Tools for ECG Data Analysis , 2006 .

[2]  A. Jovic,et al.  Feature Extraction for ECG Time-Series Mining Based on Chaos Theory , 2007, International Conference on Information Technology Interfaces.

[3]  A. Freking,et al.  Demonstration of nonlinear components in heart rate variability of healthy persons. , 1998, The American journal of physiology.

[4]  Nir Friedman,et al.  Bayesian Network Classifiers , 1997, Machine Learning.

[5]  N. Wessel,et al.  Short-term forecasting of life-threatening cardiac arrhythmias based on symbolic dynamics and finite-time growth rates. , 2000, Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics.

[6]  Madalena Costa,et al.  Multiscale entropy analysis of complex physiologic time series. , 2002, Physical review letters.