TITLE:
Individual Identification Using ECG SignalsW
AUTHORS:
Mohamad O. Diab, Alaa Seif, Mohamad El-Abed, Maher Sabbah
KEYWORDS:
Biometrics, ECG Signals, Fiducial Features, Discrete Wavelet Transform (DWT), Multilayer Perceptron (MLP)
JOURNAL NAME:
Journal of Computer and Communications,
Vol.6 No.1,
December
29,
2017
ABSTRACT: The electrocardiogram (ECG) signal used for diagnosis and patient monitoring, has recently emerged as a biometric recognition tool. Indeed, ECG signal changes from one person to another according to health status, heart geometry and anatomy among other factors. This paper forms a comparative study between different identification techniques and their performances. Previous works in this field referred to methodologies implementing either set of fiducial or set non-fiducial features. In this study we show a comparison of the same data using a fiducial feature set and a non-fiducial feature set based on statistical calculation of wavelet coefficient. High identification rates were measured in both cases, non-fiducial using Discrete Meyer (dmey) wavelet outperformed the rest at 98.65.