Identity, Privacy and Security Institute @ UofT - Communications ...

BioSec laboratory IPSI: Identity, Privacy and Security Institute @ UofT

Security concerns increase as the technology for falsification advances. There are strong evidences that a difficult to falsify biometric trait, the human heartbeat (or electrocardiogram (ECG)), can be used for identity recognition. As a clinical tool, ECG has enormous applications in heart condition diagnoses. As a biometric tool, it is particularly powerful in discriminating subjects, because of the various electrophysiological properties of the heart. Advantages and disadvantages include:

Recognition Algorithms Existing solutions for biometric recognition from electrocardiogram (ECG) signals are based on temporal and amplitude distances between detected fiducial points. Such methods rely heavily on the accuracy of fiducial detection, which is still an open problem due to the difficulty in exact localization of wave boundaries. Our systematic analysis for human identification from ECG data removes the emphasis from fiducial detection, and achieves high recognition performance with low complexity and simplicity. Our autocorrelation based method is a very simple and effective approach that does not require any waveform detection. It depends on estimating and classifying the significant coefficients of the Discrete Cosine Transform (AC/DCT) of the Linear Discriminant Analysis (AC/LDA) of the windowed autocorrelation of heartbeat signals.

Contact: D. Hatzinakos, ([email protected])

Cardiac Disorders Recently, we have extended this method to cases of arrhythmias by introducing a novel procedure for classification of healthy vs. arrhythmic ECG windows prior to ECG recognition. We designed an identification system robust to common cardiac irregularities such as premature ventricular (PVC) and atrial (APC) contractions. Criteria concerning the power distribution and complexity of the signals are used to bring to light abnormal ECG recordings, which are not employable for matching. Experimental results indicate a recognition rate of 96.2%, with misclassification taking place mostly among irregular recordings.

Privacy Concerns We have explored privacy-oriented monitoring topologies which use ECG biometric recognition. A two step framework was developed, to allocate medical data to the particular patient folders, with anonymity guarantees.

An ECG identification block acts as precursor, delivering a ranked list of potential candidates. Then a validation scheme is applied to this pruned list of candidates.

This secondary step is shown to offer more flexibility in achieving improved misclassification rates.

Time Dependency Another aspect of our research in ECG biometrics, is the investigation of destabilizing factors such as the physical or psychological activity. Physical activities may increase the heart rate, which is usually treated with HRV corrections of the features. The psychological activity on the other hand, results in very a unpredictable behavior. Since it is impossible to train the recognizers on all emotional expressions, our group has developed a methodology which can capture the instantaneous behavior of the signal in a template.


Recent Publications: Recent Publications: 1.

2.

3.

4.

5.

6.

7. 8.

9.

10. 11.

12. 13. 14.

15. 16. 17. 18.

19. 20.

21.

Majid Komeili, Wael Louis, Narges Armanfard and Dimitrios Hatzinakos, ” Feature Selection for Non-Stationary Data:Application to Human Recognition Using Medical Biometrics”, IEEE Transactions on Cybernetics, acceptred, February 2017. Wael Louis, Shahad Abdulnour, Sahar Javaher Haghighi and Dimitrios Hatzinakos,” On Biometric Systems: Electrocardiogram Gaussianity and Data Synthesis”, EURASIP Journal on Bioinformatics and bio-enginering, accepted, January 2017. Majid Komeili, Narges Armanfard and Dimitrios Hatzinakos, ” Liveness Detection and Automatic Template updating using Fusion of ECG and Fingerprint”, IEEE Transactions on Information Forensics and Security, Submitted, Dec. 2016. Wael Louis, Majid Komeili and Dimitrios Hatzinakos, ” Continuous Authentication Using One Dimensional MultiResolution Local Binary Patterns (1DMRLBP) in ECG Biometrics” , IEEE Transactions on Information Forensics and Security, Vol. 11 (12), pp. 2818-2832, December 2016. Pouryayevali, S., Wahabi, S., Hari, S. and Hatzinakos, D., ”On Evaluating ECG Biometric Systems: SessionDependency and Body Posture” IEEE Transactions on Information Forensics and Security, vol. 9(11), pp.20022012, December 2014. F. Agrafioti, F. Bui and D. Hatzinakos, ”Secure Telemedicine: Biometrics for Remote and Continuous Patient Verification”, Journal of Computer Networks and Communications, Vol 2012, Article ID 924791, 11 pages , 2012. doi:10.1155/2012/924791. F. Agrafioti, D. Hatzinakos and K. Anderson, ”ECG Pattern Analysis for Emotion Detection”, IEEE Transactins on Affective Computing, Vol. 3(1), pp. 102-115, Jan-March 2012. Francis M. Bui and Dimitris Hatzinakos, Quality of Service (QOS) Regulation in Secure Body Area Networks: System Modeling and adaptation methods, EURASIP Journal on Wireless Communications and networking Special Issue on Towards the Connected Body: Advances in Body communications, Volume 2011, Article ID 641867, 14 pages, 2011. DOI:10.1155/2011/641867 F. Agrafioti, F. M. Bui, D. Hatzinakos, Medical Biometrics in Mobile Health Monitoring, Wiley’s Security and Communication Networks Journal - Special Issue on Biometric Security for Mobile Computing, July 2010, 10.1002/sec.227. F. Agrafioti and D. Hatzinakos, ”ECG Biometric Analysis in Cardiac Irregularity Conditions”, Signal, Image and Video Processing, Springer, pp 1863-1703, September 2008. Y. Wang, F. Agrafioti, D. Hatzinakos and K.N. Plataniotis, ” Analysis of Human Electrocardiogram (ECG) for Biometric Recognition”, Hindawi Publishing Corporation EURASIP Journal on Advances in Signal Processing Volume 2008, Article ID 148658, , 11 pages, 2008 F. Agrafioti and D. Hatzinakos, ““ECG based human identification in Arrhythmia scenarios”, submitted to Eurasip Pattern Recognition, June 2007 F. M. Bui and D. Hatzinakos, " Resource allocation strategies for secure and efficient communications in biometrics-based body sensor networks", Biometrics Symposium, BSYM-2007. F. Bui and D. Hatzinakos, “ Biometric methods for secure communications in body sensor networks: Resourceefficient key management and signal-level data scrambling”, submitted to Eurasip Journal on Applied Signal Processing, May 2007. F. Agrafioti and D. Hatzinakos, “ A low complexity human identification method based on electrocardiogram (ECG) signals, to appear, Biometrics Symposium 2007. Wang, F. Agrafioti, D. Hatzinakos, and K. Plataniotis, “ Analysis of Human Electrocardiogram (ECG) for Biometric Recognition”, submitted to Eurasip Journal on Applied Signal Processing, May 2006. K. Plataniotis, D. Hatzinakos and J. Lee, “ECG Biometric Recognition without Fiducial Detection”, Biometrics Symposium/Biometrics Consortium Conference, BCC'06}, Baltimore, Sept. 19-21, 2006. Yongjin Wang, K. Plataniotis, and D. Hatzinakos, “Integrating analytic and appearance attributes for human identification from ECG signals”, Biometrics Symposium/Biometrics Consortium Conference, BCC'06}, Baltimore, Sept. 19-21, 2006. F. Agrafioti, F M. Bui and D. Hatzinakos, “On Supporting Anonymity in a BAN Biometric Framework”, Proceedings of IEEE International Conference on Digital Signal Processing DSP-2009, pp. 1-6, July 2009, Santorini, Greece F. Agrafioti, and D. Hatzinakos, “Fusion of ECG Sources for Human Identification”, Proceedings of IEEE International Symposium on Communications, Control and Signal Processing, ISCCSP-2008, pp.1542-1547 March 2008, Malta F. Agrafioti, and D. Hatzinakos, “ECG Based Recognition Using Second Order Statistics”, Proceedings of IEEE International Conference on Communication Networks and Services Research, CNSR-2008, pp. 82-87 May 2008, Halifax, Canada
