Abstract
In this paper, a neuro-fuzzy classification method is used for identifications of ECG signals. A feature extraction method with a QRS like filter (first order Gaussian derivative filter) is used. Five standard parameters (energy, mean value, standard deviation, maximum and minimum) are extracted from these diseasefeatures and then used as inputs for the neuro-fuzzy classification system. The ECG signals are importedfrom the standard MIT-BIH database. Five types of ECG signalsare used for classification; they are normal sinus rhythm (NSR), left bundle branch block (LBBB), right bundle branch block (RBBB), premature ventricular contraction (PVC) and pacemaker (PM). The proposed system combines the neural network adaptive capabilities and fuzzy inference system with the suitable filter design to give a promising classification accuracy of 99%.
