A Human ECG Identification System Based on Ensemble Empirical Mode Decomposition

• Published in: Sensors (Basel, Switzerland) Vol. 13; no. 5; pp. 6832 - 6864
• Main Authors: Zhao, Zhidong, Yang, Lei, Chen, Diandian, Luo, Yi
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 22.05.2013; Molecular Diversity Preservation International (MDPI)
• Subjects: Accuracy; Algorithms; Biometric identification; Biometrics; Discriminant analysis; ECG Identification System; Electrocardiography; ensemble empirical mode decomposition; Fourier Analysis; Heart rate; Heart Rate - physiology; Humans; Identification systems; Just noticeable difference; k-nearest neighbors; Neural networks; Physiology; Principal components analysis; Sensors; Signal Processing, Computer-Assisted; Time Factors; China
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s130506832

In this paper, a human electrocardiogram (ECG) identification system based on ensemble empirical mode decomposition (EEMD) is designed. A robust preprocessing method comprising noise elimination, heartbeat normalization and quality measurement is proposed to eliminate the effects of noise and heart rate variability. The system is independent of the heart rate. The ECG signal is decomposed into a number of intrinsic mode functions (IMFs) and Welch spectral analysis is used to extract the significant heartbeat signal features. Principal component analysis is used reduce the dimensionality of the feature space, and the K-nearest neighbors (K-NN) method is applied as the classifier tool. The proposed human ECG identification system was tested on standard MIT-BIH ECG databases: the ST change database, the long-term ST database, and the PTB database. The system achieved an identification accuracy of 95% for 90 subjects, demonstrating the effectiveness of the proposed method in terms of accuracy and robustness.