Fractal features for cardiac arrhythmias recognition using neural network based classifier

This paper proposes a method for cardiac arrhythmias recognition using fractal transformation (FT) and neural network based classifier. Iterated function system (IFS) uses the nonlinear interpolation in the map and uses similarity maps to construct various fractal features including supraventricular ectopic beat, bundle branch ectopic beat, and ventricular ectopic beat. Probabilistic neural network (PNN) is proposed to recognize normal heartbeat and multiple cardiac arrhythmias. The neural network based classifier with fractal features is tested by using the Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) arrhythmia database. The results will appear the efficiency of the proposed method, and also show high accuracy for recognizing electrocardiogram (ECG) signals.