Analysis of features for efficient ECG signal classification using neuro-fuzzy network

The paper considers the problem of optimizing the set of features following from Hermite representation of the QRS complex of the electrocardiogram signals for the classification of the heart arrhythmias. The principal component analysis as well as specially defined quality measure have been applied to verify the discriminative ability of the proposed feature set. As the classifier we have used Takagi-Sugeno-Kang neuro-fuzzy network of the modified structure and learning algorithm, well suited for large size problems. The numerical results of recognition of 7 types of different heart rhythms are presented and discussed.