Informatics Publishing Limited

Zeeshan Hassan ¹, Syed Omer Gilani ¹, Mohsin Jamil ¹

Affiliations
1 Department of Robotics and Artificial Intelligence, School of Mechanical and Manufacturing Engineering (SMME), National University of Sciences and Technology (NUST), H-12 Main Campus, Islamabad, PK

Abstract

In this paper, a non-fiducial approach using Wavelet Packet Decomposition (WPD) algorithm for repeated examination of solitary cable ECG used for individual identification is planned and tested. Multiple samples of ECG wave are extracted considering R-peak as a reference and WPD algorithm is applied for feature extraction. This feature file is fed as an input to a machine learning classifier i.e. random forest in order to classify the individuals. In this work, records from publicly available MIT/BIH arrhythmia dataset have been utilized to evaluate the proposed system. Best result relies on third level of wavelet decomposition using Daubechies wavelet to analyze the signal. Furthermore ranker search method is used in conjunction with relief attribute evaluator for feature selection and random forest classifier is applied by generating 100 trees. It is shown that the method is effective for quantifying the classification of arrhythmia ECG signals with accuracy of 92.62%.

Keywords

Biometrics, ECG, MIT-BIH Arrhythmia Database, Random Forest, Wavelet Packet Decomposition.

Full Text

   

Zeeshan Hassan ¹, Syed Omer Gilani ¹, Mohsin Jamil ¹

Affiliations
1 Department of Robotics and Artificial Intelligence, School of Mechanical and Manufacturing Engineering (SMME), National University of Sciences and Technology (NUST) H-12 Main Campus Islamabad, PK

Abstract

Background/Objectives: Biometric is a tool of measuring and statistically analyzing biological data. Due to advancement in technology, spoofing attacks and credential forgery are becoming very common issues of modern societies. Methods/ Statistical Analysis: Over the last decade, the Electrocardiogram (ECG) is known as an emerging biometric instrument for individual identification and verification as the ECG varies among people because of their diverse anatomy of the heart. At present ECG is a popular research topic in the area of physiological biometrics. The greater part of ECG biometric literature employs fiducially based features, resulting from spikes, crest and temporal marker of ECG signal. Findings: The main focus of this review is to provide scientific analysis and comparison between fiducial and non-fiducial techniques of feature extraction especially in terms of efficiency in large and small datasets. It also provides a key manifestation of future research perspectives in the field of ECG based biometrics. Application/Improvements: The proposed review can be useful in further research in the same area.

Keywords

Biometrics, ECG, Fiducial, Non-Fiducial, QRS-Complex.

Full Text