Hybrid system for cardiac arrhythmia classification with fuzzy K-nearest neighbors and neural networks combined by a fuzzy inference system

Abstract

In this paper we describe a hybrid architecture for classification of cardiac arrhythmias taking as a source the ECG records MIT-BIH Arrhythmia database. The Samples were taken from the LBBB, RBBB, PVC and Fusion Paced and Normal arrhythmias, as well as the normal heartbeats. These were segmented and transformation and 3 methods of classification were used: Fuzzy KNN, Multi Layer Perceptron with Gradient Descent and momentum Backpropagation and Multi Layer Perceptron with Scaled Conjugate Gradient Backpropagation. Finally, we used a Mamdani type fuzzy inference system to combine the outputs of each classifier, and we achieved a very high classification rate of 98%. © 2010 Springer-Verlag Berlin Heidelberg.