Ayatollahi - Iran University of Science and Technology
Jafarnia Dabanloo - Iran University of Science and Technology
McLernon - University of Leeds, Leeds, UK

Many Researchers have widely investigated developing a mathematical model for generating artificial electrocardiogram (ECG) signals. In this paper we present a new comprehensive model for artificial ECG generation. Using a new neural network approach in a nonlinear dynamical system provides the ability of generating a wide range of ECG signals. In addition, using the Integral Pulse Frequency Modulator (IPFM) model incorporates the effects of sympathetic and parasympathetic activities in simulating the heart rate variability (HRV) signals. The inter-coupling between sympathetic and parasympathetic systems is also included. In addition to modelling the abnormalities associated with the shape of ECG, this model can easily model the HRV signals associated with the autonomic regularity of heart rate sicknesses. One of the uses of this model is for easy assessment of diagnostic ECG signal processing devices.

Electrocardiogram, neural networks, dynamical systems