Radyal tabanlı fonksiyon ağlarını kullanarak EKG sinyallerinin sıkıştırılması

Abstract

Elektrokardiyogram (EKG), kalbin çalışması esnasında kalp kaslarında meydana gelen elektriksel aktivitelerin grafik olarak gösterimidir.  EKG, kalp hastalıklarının teşhisinde ve analizinde oldukça önemli bir rol oynamaktadır.  Herhangi bir kalp rahatsızlığına sahip kişilerin kalbinde meydana gelebilecek bir rahatsızlığı önceden tespit edebilmek için, EKG sinyalleri sürekli olarak kaydedilir, depolanır ve dijital iletişim ağları üzerinden iletilir.  Ancak bu tür kayıtlar ortamdan dolayı gürültüye maruz kalabilir.  Dahası, bu şekildeki kayıtlar depolama ve iletimi zorlaştıracak düzeyde büyük miktarda veri üretir.  Yukarıda sözü edilen nedenlerden dolayı gürültülü ortamda bile etkili bir EKG veri sıkıştırma modeli gereklidir.  Bu çalışma, EKG işaretlerinin doğal yapısını gürültülü ortamlarda bile korumak ve daha az sayıda parametre ile yeniden temsil etmek için Radyal Tabanlı Fonksiyon Ağlarını (RTFA) sunar.  RTFA’ların tasarımında, modelin yaklaşık doğruluğunu etkileyen önemli unsurlardan birisi olan radyal taban fonksiyonlarının merkezlerinin verimli bir şekilde belirlenmesidir.  Bu amaçla, k-means kümeleme algoritması kullanılmıştır. Yeniden yapılandırılmış EKG dalga biçimi, ortalama karesel hata, ortalama mutlak hata ve sıkıştırma oranı açısından niceliksel olarak değerlendirilmiştir. Tüm bu adımlar MATLAB ortamında uygulanmıştır.

Keywords

• Elektrokardiogram,radyal tabanlı fonksiyon ağları,veri sıkıştırma

ECG signal compression using radial basis function networks

Abstract

An electrocardiogram (ECG) is the graphical representation of electrical activity in the cardiac muscles of the heart.  It plays a significant role in diagnosis and analysis of cardiac diseases.  In order to detect any cardiac diseases in advance, the ECG signals are continuously recorded, stored and transmitted over digital communication networks, but such records may be subject to noise due to environment.  Moreover, these types of records produce large amounts of data that will make storage and transmission difficult.  Due to the reasons mentioned above, an effective ECG data compression model is required even in a noisy environment.  This work presents Radial Basis Function Networks (RBFN) to preserve the natural structure of ECG signals even in noisy environments and to re-construct with fewer parameters.  In the design of RBFN, the center of the radial basis functions, which is one of the important factors affecting the approximate accuracy of the model, is to be determined efficiently.  For this purpose, k-means clustering algorithm is used in the paper.  The reconstructed ECG waveform was quantitatively evaluated in terms of root mean squared error, mean absolute error, and compression ratio. These steps are implemented in MATLAB environment.

Keywords

• Electrocardiogram,radial basis function networks,compression

References

1. Jalaleddine, S. M., Hutchens, C. G., Strattan, R. D. ve Coberly, W. A.., ECG data compression techniques-a unified approach, IEEE Transsctions on Biomedical Engineering, 37(4), 329-343, (1990).
2. Ishijima, M., Shin, S. B., Hostetter, G. H. ve Sklansky, J., Scan-along polygonal approximation for data compression of electrocardiograms, IEEE Transactions on Biomedical Engineering, 11, 723-729, (1983).
3. Horspool, R. N. ve Windels, W. J., ECG compression using Ziv-Lempel techniques, Computers and biomedical research, 28(1), 67-86, (1995).
4. Imai, H., Kiraura, N. ve Yoshlda, Y, An efficient encoding method for electrocardiography using spline functions, Systems and Computers in Japan, 16(3), 85-94, (1985).
5. Barlas, G. D. ve Skordalakis, E. S., A novel family of compression algorithms for ECG and other semiperiodical, one-dimensional, biomedical signals, IEEE transactions on biomedical engineering, 43(8), 820-828, (1996).
6. Reddy, B. S. ve Murthy, I. S. N., ECG data compression using Fourier descriptors, IEEE Transactions on Biomedical Engineering, 4, 428-434, (1986).
7. Al-Nashash, H. A. M., ECG data compression using adaptive Fourier coefficients estimation, Medical engineering & physics, 16, 1, 62-66, (1994).
8. Benzid, R., Messaoudi, A. ve Boussaad, A., Constrained ECG compression algorithm using the block-based discrete cosine transform, Digital Signal Processing,18(1), 56-64, (2008).

9. Bendifallah, A., Benzid, R. ve Boulemden, M., Improved ECG compression method using discrete cosine transform. Electronics letters, 47(2), 87-89, (2011).
10. Chen, J., Itoh, S. ve Hashimoto, T., ECG data compression by using wavelet transform, IEICE Transactions on Information and Systems, 76, 12, 1454-1461, (1993).
11. Patel, S. ve Datar, A., ECG data compression using wavelet transform. International Journal of Engineering Trends & Technology, 10, 770-776, (2014).
12. Manikandan, M. S. ve Dandapat, S., Wavelet-based electrocardiogram signal compression methods and their performances: a prospective review, Biomedical Signal Processing and Control, 14, 73-107, (2014).
13. Addison, P. S., Wavelet transforms and the ECG: a review, Physiological Measurement, 26(5), R155, (2005).
14. Abo-Zahhad, M., Ahmed, S. M., Sabor, N. ve Al-Ajlouni, A. F., Wavelet threshold based ECG data compression technique using immune optimization algorithm, International Journal of Signal Processing, Image Processing and Pattern Recognition, 8(2), 307-360, (2015).
15. Swarnkar, A., Kumar, R., Kumar, A. ve Khanna, P., Performance of different threshold function for ECG compression using Slantlet transform, Proceedings, 4th International Conference on Signal Processing and Integrated Networks, 375-379, Noida, India, (2017).
16. Ballesteros, D. M., Moreno, D. M. ve Gaona, A. E., FPGA compression of ECG signals by using modified convolution scheme of the Discrete Wavelet Transform, Ingeniare, Revista chilena de ingeniería, 20, 1, (2012).
17. Al-Busaidi, A. M., Khriji, L., Touati, F., Rasid, M. F. A. ve Mnaouer, A. B., Real-time DWT-based compression for wearable electrocardiogram monitoring system, Proceedings, IEEE 8th GCC Conference and Exhibition (GCCC), 1-6, Muscat, Umman, (2015).
18. Huang, B., Wang, Y. ve Chen, J., ECG compression using the context modeling arithmetic coding with dynamic learning vector–scalar quantization, Biomedical Signal Processing and Control, 8(1), 59-65, (2013).
19. Hung, K. C., Wu, T. C., Lee, H. W. ve Liu, T. K., EP-based wavelet coefficient quantization for linear distortion ECG data compression, Medical Engineering & Physics, 36(7), 809-821, (2014).
20. Ramakrishnan, A. G. ve Saha, S., ECG coding by wavelet-based linear prediction, IEEE Transactions on Biomedical Engineering, 44, 12, 1253-1261, (1997).
21. Al-Shrouf, A., Abo-Zahhad, M. ve Ahmed, S. M., A novel compression algorithm for electrocardiogram signals based on the linear prediction of the wavelet coefficients, Digital Signal Processing, 13, 4, 604-622, (2003).
22. Le Gia, Q. T. ve Wendland, H., Data compression on the sphere using multiscale radial basis function approximation, Advances in Computational Mathematics, 40(4), 923-943, (2014).
23. Balasubramani, P. ve Murugan, P. R. Efficient image compression techniques for compressing multimodal medical images using neural network radial basis function approach. International Journal of Imaging Systems and Technology, 25(2), 115-122, (2015).
24. Perumal, B., Rajasekaran, M. P. ve Murugan, H., Comparison of neural network algorithms in image compression technique. Proceedings, 3rd International Conference on Emerging Technological Trends (ICETT), Kerala, India, 1-6, (2016).
25. Jasmi, R. P., Perumal, B. ve Rajasekaran, M. P., Comparison of medical image compression using DWT algorithm and neural network techniques. Advances in Natural and Applied Sciences, 8, 19, 1-10, (2014).
26. Park, J. ve Sandberg, I. W., Universal approximation using radial-basis-function networks, Neural computation, 3(2), 246-257, (1991).
27. Ranjeet, K., Kumar, A. ve Pandey, R. K., ECG signal compression using different techniques. Advances in Computing, Communication and Control, 231-241, (2011).