Machine Learning-Empowered Biometric Methods for Biomedicine Applications

Qingxue Zhang; Dian Zhou; Xuan Zeng; Qingxue Zhang; Dian Zhou; Xuan Zeng

doi:10.3934/medsci.2017.3.274

AIMS Medical Science

2017, Volume 4, Issue 3: 274-290. doi: 10.3934/medsci.2017.3.274

Previous Article Next Article

Research article Special Issues

Machine Learning-Empowered Biometric Methods for Biomedicine Applications

1.
Department of Electrical Engineering, University of Texas at Dallas, 800 W Campbell Rd, Richardson, TX 75080, USA;
2.
Department of Microelectronics, Fudan University, 220 Handan Rd, Shanghai, 200433, China

Received: 29 March 2017 Accepted: 21 June 2017 Published: 20 July 2017

Nowadays, pervasive computing technologies are paving a promising way for advanced smart health applications. However, a key impediment faced by wide deployment of these assistive smart devices, is the increasing privacy and security issue, such as how to protect access to sensitive patient data in the health record. Focusing on this challenge, biometrics are attracting intense attention in terms of effective user identification to enable confidential health applications. In this paper, we take special interest in two bio-potential-based biometric modalities, electrocardiogram (ECG) and electroencephalogram (EEG), considering that they are both unique to individuals, and more reliable than token (identity card) and knowledge-based (username/password) methods. After extracting effective features in multiple domains from ECG/EEG signals, several advanced machine learning algorithms are introduced to perform the user identification task, including Neural Network, K-nearest Neighbor, Bagging, Random Forest and AdaBoost. Experimental results on two public ECG and EEG datasets show that ECG is a more robust biometric modality compared to EEG, leveraging a higher signal to noise ratio and also more distinguishable morphological patterns. Among different machine learning classifiers, the random forest greatly outperforms the others and owns an identification rate as high as 98%. This study is expected to demonstrate that properly selected biometric empowered by an effective machine learner owns a great potential, to enable confidential biomedicine applications in the era of smart digital health.
- biomedicine,
- smart health,
- biometrics,
- user identification,
- electrocardiogram,
- electroencephalogram,
- neural network,
- random forest,
- AdaBoost
Citation: Qingxue Zhang, Dian Zhou, Xuan Zeng. Machine Learning-Empowered Biometric Methods for Biomedicine Applications[J]. AIMS Medical Science, 2017, 4(3): 274-290. doi: 10.3934/medsci.2017.3.274

Related Papers:

Abstract

Nowadays, pervasive computing technologies are paving a promising way for advanced smart health applications. However, a key impediment faced by wide deployment of these assistive smart devices, is the increasing privacy and security issue, such as how to protect access to sensitive patient data in the health record. Focusing on this challenge, biometrics are attracting intense attention in terms of effective user identification to enable confidential health applications. In this paper, we take special interest in two bio-potential-based biometric modalities, electrocardiogram (ECG) and electroencephalogram (EEG), considering that they are both unique to individuals, and more reliable than token (identity card) and knowledge-based (username/password) methods. After extracting effective features in multiple domains from ECG/EEG signals, several advanced machine learning algorithms are introduced to perform the user identification task, including Neural Network, K-nearest Neighbor, Bagging, Random Forest and AdaBoost. Experimental results on two public ECG and EEG datasets show that ECG is a more robust biometric modality compared to EEG, leveraging a higher signal to noise ratio and also more distinguishable morphological patterns. Among different machine learning classifiers, the random forest greatly outperforms the others and owns an identification rate as high as 98%. This study is expected to demonstrate that properly selected biometric empowered by an effective machine learner owns a great potential, to enable confidential biomedicine applications in the era of smart digital health.

References

[1]	Lymberis A (2003) Smart wearable systems for personalised health management: current R&D and future challenges. In: Engineering in Medicine and Biology Society. Proceedings of the 25th Annual International Conference of the IEEE. IEEE 4: 3716-3719.
[2]	Solanas A, Patsakis C, Conti M, et al. (2014) Smart health: a context-aware health paradigm within smart cities. IEEE Communications Magazine 52: 74-81.
[3]	Birjandtalab J, Zhang Q, Jafari R (2015) A case study on minimum energy operation for dynamic time warping signal processing in wearable computers. In: Pervasive Computing and Communication Workshops (PerCom Workshops), 2015 IEEE International Conference: 415-420.
[4]	Eriksson J, Vaka P, Shen F, et al. PerMoby'15: The Fourth IEEE International Workshop on the Impact of Human Mobility in Pervasive Systems and Applications, 2015-Program.
[5]	Zhang Q, Zhou D, Zeng X (2017) Highly wearable cuff-less blood pressure and heart rate monitoring with single-arm electrocardiogram and photoplethysmogram signals. Biomed Eng Online 16: 23. doi: 10.1186/s12938-017-0317-z
[6]	Zhang Q, Zahed C, Nathan V, et al. (2015) An ECG dataset representing real-world signal characteristics for wearable computers. In: Biomedical Circuits and Systems Conference (BioCAS), 2015 IEEE. IEEE: 1-4.
[7]	Zhang Q, Zhou D, Zeng X (2016) A Novel Framework for Motion-tolerant Instantaneous Heart Rate Estimation By Phase-domain Multi-view Dynamic Time Warping. IEEE Transactions on Biomedical Engineering.
[8]	Zhang Q, Zhou D, Zeng X (2016) A novel machine learning-enabled framework for instantaneous heart rate monitoring from motion-artifact-corrupted electrocardiogram signals. Physiol Meas 37: 1945-1967. doi: 10.1088/0967-3334/37/11/1945
[9]	Unar JA, Seng WC, Abbasi A (2014) A review of biometric technology along with trends and prospects. Pattern Recogn 47: 2673-2688.
[10]	Yao J, Wan Y (2008) A wavelet method for biometric identification using wearable ECG sensors. In: Medical Devices and Biosensors, 2008. ISSS-MDBS 2008. 5th International Summer School and Symposium on. IEEE: 297-300.
[11]	Paranjape RB, Mahovsky J, Benedicenti L, et al. (2001) The electroencephalogram as a biometric. In: Electrical and Computer Engineering, 2001. Canadian Conference on. IEEE 2: 1363-1366. doi: 10.1109/CCECE.2001.933649
[12]	Shalev-Shwartz S, Ben-David S (2014) Understanding machine learning: From theory to algorithms. Cambridge university press.
[13]	Iyengar N, Peng CK, Morin R, et al. (1996) Age-related alterations in the fractal scaling of cardiac interbeat interval dynamics. Am J Physiol-Reg I 71: R1078-R1084.
[14]	Goldberger AL, Amaral L, Glass L, et al. (2000) Physiobank, physiotoolkit, and physionet. Circulation. Discovery 101: e215-e220. doi: 10.1161/01.CIR.101.23.e215
[15]	Begleiter H. UCI EEG Database. Available from: https://archive.ics.uci.edu/ml/datasets/EEG+Database.
[16]	Bai J, Ng S (2005) Tests for skewness, kurtosis, and normality for time series data, J Bus Econ Stat 23: 49-60.
[17]	Phinyomark A, Limsakul C, Phukpattaranont P (2009) A novel feature extraction for robust EMG pattern recognition. arXiv preprint arXiv:0912.3973, 2009.
[18]	Palaniappan R (2004) Method of identifying individuals using VEP signals and neural network. IEE P-SCI Meas Tech 151: 16-20. doi: 10.1049/ip-smt:20040003
[19]	F. f. O. A. Statistics. RStudio, Available from: https://www.rstudio.com/.
[20]	Machine learning: An artificial intelligence approach. Springer Science & Business Media, 2013.
[21]	Tang X, Shu L (2014) Classification of electrocardiogram signals with RS and quantum neural networks. IJMUE 9:363-372.
[22]	Lourenço A, Silva H, Fred A (2011) Unveiling the biometric potential of finger-based ECG signals. Comput Intell Neurosci 2011: 5.
[23]	Ting CM, Salleh SH (2010) ECG based personal identification using extended kalman filter. In: Information Sciences Signal Processing and their Applications (ISSPA), 2010 10th International Conference on. IEEE: 774-777.
[24]	Aboalayon KAI, Faezipour M, Almuhammadi WS, et al. (2016) Sleep Stage Classification Using EEG Signal Analysis: A Comprehensive Survey and New Investigation. Entropy 18: 272. doi: 10.3390/e18090272
[25]	Hassan AR, Bhuiyan MIH (2016). A decision support system for automatic sleep staging from EEG signals using tunable q-factor wavelet transform and spectral features. J Neurosci Methods 271: 107-118. doi: 10.1016/j.jneumeth.2016.07.012
[26]	Hassan AR, Subasi A (2016) Automatic identification of epileptic seizures from EEG signals using linear programming boosting. Comput Meth Prog Bio 136: 65-77. doi: 10.1016/j.cmpb.2016.08.013
[27]	Hassan AR, Haque MA (2017) An expert system for automated identification of obstructive sleep apnea from single-lead ECG using random under sampling boosting. Neurocomputing 235: 122-130. doi: 10.1016/j.neucom.2016.12.062
[28]	Hassan AR (2016) Computer-aided obstructive sleep apnea detection using normal inverse Gaussian parameters and adaptive boosting. Biomed Signal Proces 29: 22-30.

Reader Comments

Your name:*

Email:*
© 2017 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)