Human Identification by Cross-Correlation and Pattern Matching of Personalized Heartbeat: Influence of ECG Leads and Reference Database Size

Jekova, Irena; Krasteva, Vessela; Schmid, Ramun

doi:10.3390/s18020372

Cited by 13 publications

(12 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the first attempt of using ECG for biometric purposes was presented by Biel et al in 1999 and [23,24]. Since then the researchers devoted much of their efforts to utilizing different methodologies to improve the ECG-based biometric performance [1,7,8,[25][26][27]. Different databases, both private and public, are used to evaluate the proposed approaches [28].…”

Section: Introductionmentioning

confidence: 99%

ECG-Based Subject Identification Using Common Spatial Pattern and SVM

et al. 2019

View full text Add to dashboard Cite

In this paper, a nonfiducial electrocardiogram (ECG, the process of recording the electrical activity of the heart over a period of time using electrodes placed on the skin) identification system based on the common spatial pattern (CSP) feature extraction technique is presented. The single- and multilead ECG signals of each subject are divided into nonoverlapping segments, and different segment lengths (1, 3, 5, 7, 10, or 15 seconds) are investigated. Features are extracted from each signal segment through projection on a CSP projection matrix. The extracted features are then used to train a radial basis function kernel-based Support Vector Machine (SVM) classifier, which is then employed in the identification phase. The proposed identification system was evaluated on 10, 20, …, 200 reference subjects of the Physikalisch-Technische Bundesanstalt (PTB) ECG database. Using a single limb-based lead (I) with 200 reference subjects, the system achieved an identification rate of 95.15% and equal error rate of 0.1. The use of a single chest-based lead (V3) for 200 reference subjects resulted in an identification rate of 98.92% and equal error rate of 0.08.

show abstract

Section: Introductionmentioning

confidence: 99%

ECG-Based Subject Identification Using Common Spatial Pattern and SVM

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Wübbeler et al [ 24 ] showed a method based on the distance between the first and second temporal derivatives of signals. Poree et al [ 25 ] used the maximal correlation coefficient applied over a 12-lead ECG, and a cross correlation of a 12-lead ECG was used for human verification [ 26 ] and identification [ 27 ]. Many researchers have suggested various sophisticated algorithms to classify the ECG signals, e.g., neural networks [ 20 , 28 , 29 ], independent component analysis [ 30 ], k-nearest neighborhood [ 31 , 32 , 33 , 34 ], and support vector machine [ 35 , 36 , 37 , 38 , 39 ].…”

Section: Introductionmentioning

confidence: 99%

Individual Biometric Identification Using Multi-Cycle Electrocardiographic Waveform Patterns

Lee

Kim

2018

Sensors

View full text Add to dashboard Cite

The electrocardiogram (ECG) waveform conveys information regarding the electrical property of the heart. The patterns vary depending on the individual heart characteristics. ECG features can be potentially used for biometric recognition. This study presents a new method using the entire ECG waveform pattern for matching and demonstrates that the approach can potentially be employed for individual biometric identification. Multi-cycle ECG signals were assessed using an ECG measuring circuit, and three electrodes can be patched on the wrists or fingers for considering various measurements. For biometric identification, our-fold cross validation was used in the experiments for assessing how the results of a statistical analysis will generalize to an independent data set. Four different pattern matching algorithms, i.e., cosine similarity, cross correlation, city block distance, and Euclidean distances, were tested to compare the individual identification performances with a single channel of ECG signal (3-wire ECG). To evaluate the pattern matching for biometric identification, the ECG recordings for each subject were partitioned into training and test set. The suggested method obtained a maximum performance of 89.9% accuracy with two heartbeats of ECG signals measured on the wrist and 93.3% accuracy with three heartbeats for 55 subjects. The performance rate with ECG signals measured on the fingers improved up to 99.3% with two heartbeats and 100% with three heartbeats of signals for 20 subjects.

show abstract

“…Biel et al's [12,13] works are considered the first attempt to use ECGs for biometric purposes, considering the biometric characteristics of measurability (ease with which the characteristic is obtained), permanence (no change over time), universality (possession of the characteristic by the individual), and uniqueness (no two individuals share the same characteristic) [14][15][16][17]. Since then, many researchers have proposed various ECG-based identification approaches [1,4,[18][19][20][21][22][23][24][25][26][27] using private and/or public databases [28,29].…”

Section: Introductionmentioning

confidence: 99%

ECG-Based Subject Identification Using Statistical Features and Random Forest

et al. 2019

View full text Add to dashboard Cite

In this work, a nonfiducial electrocardiogram (ECG) identification algorithm based on statistical features and random forest classifier is presented. Two feature extraction approaches are investigated: direct and band-based approaches. In the former, eleven simple statistical features are directly extracted from a single-lead ECG signal segment. In the latter, the single-lead ECG signal is first decomposed into bands, and the statistical features are extracted from each segment of a given band and concatenated to form the feature vector. Nonoverlapping segments of different lengths (i.e., 1, 3, 5, 7, 10, or 15 sec) are examined. The extracted feature vectors are applied to a random forest classifier, for the purpose of identification. This study considers 290 reference subjects from the ECG database of the Physikalisch-Technische Bundesanstalt (PTB). The proposed identification algorithm achieved an accuracy rate of 99.61% utilizing the single limb lead (I) with the band-based approach. A single chest lead (V1), augmented limb lead (aVF), and Frank’s lead (Vx) achieved an accuracy rate of 99.37%, 99.76%, and 99.76%, respectively, using the same approach.

show abstract

Human Identification by Cross-Correlation and Pattern Matching of Personalized Heartbeat: Influence of ECG Leads and Reference Database Size

Cited by 13 publications

References 28 publications

ECG-Based Subject Identification Using Common Spatial Pattern and SVM

ECG-Based Subject Identification Using Common Spatial Pattern and SVM

Individual Biometric Identification Using Multi-Cycle Electrocardiographic Waveform Patterns

ECG-Based Subject Identification Using Statistical Features and Random Forest

Contact Info

Product

Resources

About