Documente Academic
Documente Profesional
Documente Cultură
Pratik
Roll No: 14EC291
Supervisor
Dr. Gayadhar Pradhan
• Introduction
• Challenges in ECG Signal Processing
• Past Research Work
• Current Research Work
• Future Research Work
• Publications
• References
ECG signal
• The heart is an electrical organ, and its activity can be measured
non-invasively [1].
• ECG signal provides an electrical picture of the heart and information about
different pathological conditions.
• Arrhythmia is any irregular heartbeat.
Ambulatory ECG
Why it is done
• Look for and record irregular heartbeats that come and go or happen during
certain activities.
• Find out what is causing chest pain or pressure, dizziness, or fainting. These
may be symptoms of heart problems.
• Check to see if treatment for an irregular heartbeat is working.
• During ECG recording process it gets corrupted by various noise sources [1].
These noise sources distort the morphology of the signal
• For proper Arrhythmia Classification the original signal structure must be
retrieved from the noisy record.
• In case of Ambulatory ECG monitoring the ECG gets corrupted by additive
Gaussian Noise. [3, 4, 5]
• ECG feature Extraction and Feature Selection for Proper Arrhythmia
classification [6, 7, 8]
• Researches going on in Arrhythmia Detection and classification [9, 10, 7, 11]
• ECG beat segmentation is another field to work on. [12, 13]
• Low data rate transmission during wireless ambulatory ECG
monitoring. [14, 15]
AWGN
Noise
Modified
Nonlocal Mean
Empirical Mode
Denoising
Decomposition
Soft-Thresholding
IHP for Noise
on 1st Three
Detection
IMF
Optimal Threshold
Denoised Output
condition check
Figure 3 : Block diagram representation of the proposed method for ECG denoising.
Figure 4 : Illustrating the differences during the calculation of IPs in the standard (left) and the modified EMD (right).
Experimental Results
1
0 (a)
-1
100 200 300 400 500 600 700 800
1
0 (b)
-1
100 200 300 400 500 600 700 800
Amplitude (mv)
1
0 (c)
-1
100 200 300 400 500 600 700 800
1
0 (d)
-1
100 200 300 400 500 600 700 800
1
0 (e)
-1
100 200 300 400 500 600 700 800
1
0 (f)
-1
100 200 300 400 500 600 700 800
samples
Figure 5 : Denoised output of different denoising algorithm (a) clean ECG signal, (b) noisy ECG signal, (c) Wavelet soft-thresholding, (d) EMD
soft-thresholding, (e) NLM method and (f) proposed method output for a qualitative comparison. For better visualization, 700 samples are taken for this
analysis.
Figure 6 : Comparison of the SNR improvement for the Figure 7 : Comparison of improvement in SNR with
different denoising algorithm studied in this work for varying input SNR levels for the different denoising
various ECG records. methods explored in this study.
Table 1 : The MSE values for the explored approaches and the proposed method with respect to all test
signals at fixed SNR value of 10 dB
Figure 10 : Comparison of MSE values for the different denoising algorithms with varying input
SNR levels.
Conclusion
• The problem of under-averaged samples in the case of NLM-based ECG
denoising is addressed.
• To reduce the computational cost, the modified-EMD (M-EMD) is used in
place of standard EMD method.
• Proposed approach are observed to be much superior to those for the
state-of-the-art methods.
• Technique can be explored for various other kind of noise.
• Improvements are still possible in NLM for over-smoothing problems.
Experimental Results
The result for SNR improvement , mean square error (MSE) and percent root
Distortion (PRD) is shown in the tables below:
Table 2 : The SNR improvement values of the explored approaches and the proposed method with respect to
all test signals at fixed SNR value of 10 dB
Table 3 : The MSE values of the explored approaches and the proposed method with respect to all test
signals at fixed SNR value of 10 dB
Table 4 : The PRD values of the explored approaches and the proposed method with respect to all test
signals at fixed SNR value of 10 dB
Original ECG
1
-1
0 500 1000 1500 2000
Amplitude (mv)
Noisy ECG
1
-1
0 500 1000 1500 2000
Denoised Output
1
-1
0 500 1000 1500 2000
samples [n]
Figure 13 : Denoising result of Proposed method
Conclusion
• overcomes the individual issues of both technique like rare-patch effect in
NLM and large decomposition levels in DWT based denoising methods.
• To reduce the computational cost, the proposed method uses only two level
of DWT based decomposition.
• Proposed approach are observed to be much superior to those for the
state-of-the-art methods.
LD, Linear Discriminant; VCG, Vectorcardiogram; SFFS, sequential forward floating search; CRF, conditional random fields; HBF, Hermite Basis Function;
HOS, Higher order Statistics; ICA, Independent Component analysis; PCA, Principal component analysis.
Toolkits
• Journal Article
• Pratik Singh, Gayadhar Pradhan and S Shahnawazuddin, “Denoising of ECG
signal by non-local estimation of approximation coefficients in DWT.
Biocybernetics and Biomedical Engineering,” 37(3), pp.599-610, 2017.
• Conference
• Pratik Singh, S Shahnawazuddin, Gayadhar Pradhan, “ SIGNIFICANCE OF
MODIFIED EMPIRICAL MODE DECOMPOSITION FOR ECG DENOISING”,
in IEEE Engineering in Medicine and Biological Society (EMBC), IEEE,
2017,July, pp. 2956-2959.
• Ashish Verma, Pratik Singh, Gayadhar Pradhan, “Electrocardiogram denoising
using Wavelet decomposition and EMD domain filtering.” InRegion 10
Conference (TENCON), 2016 IEEE 2016, Nov 22, pp. 2185-2189.
• Journal Communicated
• Pratik Singh, S Shahnawazuddin and Gayadhar Pradhan, “An Efficient ECG
Denoising Technique based on Non-local Means Estimation and Modified
Empirical Mode Decomposition”.(Springer)
• Conference Communicated
• Pratik Singh, Sarmila Garnaik, Gayadhar Pradhan, “Significance of non-local
means estimation in DWT based ECG signal denoising” , TENCON(2017).
References
[1] G. D. Clifford, F. Azuaje, and P. McSharry, Advanced methods and tools for ECG data analysis. Artech house London, 2006.
[2] A. Alesanco and J. Garcı́a, “Clinical assessment of wireless ECG transmission ı́n real-time cardiac telemonitoring,” IEEE Transactions on Information
Technology in Biomedicine, vol. 14, no. 5, pp. 1144–1152, Sept 2010.
[3] D. L. Donoho, “De-noising by soft-thresholding,” IEEE transactions on information theory, vol. 41, no. 3, pp. 613–627, 1995.
[4] B. H. Tracey and E. L. Miller, “Nonlocal means denoising of ECG signals,” IEEE Transactions on Biomedical Engineering, vol. 59, no. 9, pp.
2383–2386, 2012.
[5] M. A. Kabir and C. Shahnaz, “Denoising of ecg signals based on noise reduction algorithms in emd and wavelet domains,” Biomedical Signal
Processing and Control, vol. 7, no. 5, pp. 481–489, 2012.
[6] C. Alexakis, H. Nyongesa, R. Saatchi, N. Harris, C. Davies, C. Emery, R. Ireland, and S. Heller, “Feature extraction and classification of
electrocardiogram (ECG) signals related to hypoglycaemia,” Computers in Cardiology, 2003, pp. 537–540, 2003.
[7] G. Aslantas, F. Gurgen, and A. A. Salah, “GA-NN approach for ECG feature selection in rule based arrhythmia classification,” Neural Network
World, vol. 24, no. 3, pp. 267–283, 2014.
[8] S. Shilaskar and A. Ghatol, “Expert Systems with Applications Feature selection for medical diagnosis : Evaluation for cardiovascular diseases,”
Expert Systems With Applications, vol. 40, no. 10, pp. 4146–4153, 2013. [Online]. Available: http://dx.doi.org/10.1016/j.eswa.2013.01.032
[9] E. José, S. Luz, W. Robson, G. Cámara-chávez, and D. Menotti, “ECG-based heartbeat classification for arrhythmia detection : A survey,”
Computer Methods and Programs in Biomedicine, vol. 127, pp. 144–164, 2015. [Online]. Available: http://dx.doi.org/10.1016/j.cmpb.2015.12.008
[10] J.-S. Wang, W.-C. Chiang, Y.-L. Hsu, and Y.-T. C. Yang, “ECG arrhythmia classification using a probabilistic neural network with a feature reduction
method,” Neurocomputing, vol. 116, pp. 38–45, 2013. [Online]. Available: http://www.sciencedirect.com/science/article/pii/S0925231212006960
[11] G. K. Prasad and J. S. Sahambi, “ECG Arrhythmias using Multi-Resolution Analysis and Neural Networks,” TENCON 2003. Conference on
Convergent Technologies for Asia-Pacific Region, no. 3, pp. 227 – 231, 2003. [Online]. Available:
http://ieeexplore.ieee.org/xpls/abs{ }all.jsp?arnumber=1273320{&}tag=1
[12] R. V. Andreão and J. Boudy, “Combining Wavelet Transform and Hidden Markov Models for ECG Segmentation,” EURASIP Journal on Advances in
Signal Processing, vol. 2007, no. 1, pp. 1–9, 2007. [Online]. Available: http://www.hindawi.com/journals/asp/2007/056215.abs.html
[13] S. Kurtek and W. Wu, “Segmentation, alignment and statistical analysis of biosignals with application to disease classification,” Journal of Applied
. . . , vol. 40, no. 6, pp. 37–41, 2013. [Online]. Available: http://www.tandfonline.com/doi/abs/10.1080/02664763.2013.785492
References (contd.)
[14] H. Kim, R. F. Yazicioglu, P. Merken, C. Van Hoof, and H. J. Yoo, “ECG signal compression and classification algorithm with quad level vector for
ECG holter system,” IEEE Transactions on Information Technology in Biomedicine, vol. 14, no. 1, pp. 93–100, 2010.
[15] S. Padhy, L. N. Sharma, and S. Dandapat, “Multilead ECG data compression using SVD in multiresolution domain,” Biomedical Signal Processing
and Control, vol. 23, pp. 10–18, 2016. [Online]. Available: http://dx.doi.org/10.1016/j.bspc.2015.06.012
[16] P. Singh, S. Shahnawazuddin, and G. Pradhan, “Significance of modified empirical mode decomposition for ecg denoising,” in Engineering in
Medicine and Biological Society (EMBC), 39th International Conference. IEEE, 2017, pp. 2956–2959.
[17] P. Singh, G. Pradhan, and S. Shahnawazuddin, “Denoising of ecg signal by non-local estimation of approximation coefficients in dwt,”
Biocybernetics and Biomedical Engineering, vol. 37, no. 3, pp. 599–610, 2017.
[18] R. Sharma and S. M. Prasanna, “A better decomposition of speech obtained using modified empirical mode decomposition,” Digital Signal
Processing, vol. 58, pp. 26 – 39, 2016. [Online]. Available: http://www.sciencedirect.com/science/article/pii/S1051200416300975
[19] P. De Chazal, M. O’Dwyer, and R. B. Reilly, “Automatic classification of heartbeats using ecg morphology and heartbeat interval features,” IEEE
Transactions on Biomedical Engineering, vol. 51, no. 7, pp. 1196–1206, 2004.
[20] M. Llamedo and J. P. Martı́nez, “Heartbeat classification using feature selection driven by database generalization criteria,” IEEE Transactions on
Biomedical Engineering, vol. 58, no. 3, pp. 616–625, 2011.
[21] Z. Zhang, J. Dong, X. Luo, K.-S. Choi, and X. Wu, “Heartbeat classification using disease-specific feature selection,” Computers in biology and
medicine, vol. 46, pp. 79–89, 2014.
[22] M. A. Escalona-Morán, M. C. Soriano, I. Fischer, and C. R. Mirasso, “Electrocardiogram classification using reservoir computing with logistic
regression,” IEEE journal of biomedical and health informatics, vol. 19, no. 3, pp. 892–898, 2015.
[23] H. Huang, J. Liu, Q. Zhu, R. Wang, and G. Hu, “A new hierarchical method for inter-patient heartbeat classification using random projections and rr
intervals,” Biomedical engineering online, vol. 13, no. 1, p. 90, 2014.
[24] C.-C. Lin and C.-M. Yang, “Heartbeat classification using normalized rr intervals and morphological features,” Mathematical Problems in
Engineering, vol. 2014, 2014.
[25] G. De Lannoy, D. François, J. Delbeke, and M. Verleysen, “Weighted svms and feature relevance assessment in supervised heart beat classification,”
in International Joint Conference on Biomedical Engineering Systems and Technologies. Springer, 2010, pp. 212–223.
References (contd.)
[26] C. Ye, B. V. Kumar, and M. T. Coimbra, “Combining general multi-class and specific two-class classifiers for improved customized ecg heartbeat
classification,” in Pattern Recognition (ICPR), 2012 21st International Conference on. IEEE, 2012, pp. 2428–2431.
[27] Y. Bazi, N. Alajlan, H. AlHichri, and S. Malek, “Domain adaptation methods for ecg classification,” in Computer Medical Applications (ICCMA),
2013 International Conference on. IEEE, 2013, pp. 1–4.
[28] K. Dragomiretskiy and D. Zosso, “Variational mode decomposition,” IEEE transactions on signal processing, vol. 62, no. 3, pp. 531–544, 2014.
Thank you