|
Vol: 51(65) No: 4 / December 2006 Unified Neural Network-based Adaptive ECG Signal Analysis and Compression Sandor M. Szilagyi Faculty of Technical and Human Science, Sapientia – Hungarian Science University of Transylvania, Targu Mures, Romania, 540053 Targu Mures, P-ta Trandafirilor nr. 61, phone: (40) 265-206210, e-mail: szs@ms.sapientia.ro Laszlo Szilagyi Faculty of Technical and Human Science, Sapientia – Hungarian Science University of Transylvania, Targu Mures, Romania, 540053 Targu Mures, P-ta Trandafirilor nr. 61, phone: (40) 265-206210, e-mail: lalo@ms.sapientia.ro David Iclanzan Faculty of Mathematics and Computer Science, University of Babes-Bolyai, 400084 Str. M. Kogalniceanu nr. 1, Cluj Napoca, Romania, phone: (40) 264-405300/5240, e-mail: iclanzand@yahoo.com Zoltan Benyo Department of Control Engineering and Information Technology, Budapest University of Technology and Economics, Magyar Tudosok krt. 2, 1117 Budapest, Hungary, phone: (36) 1-463-1410, e-mail: benyo@iit.bme.hu Keywords: medical applications, real-time systems, unified neural network, support vector machine, adaptive signal processing, biological models. Abstract This paper presents an adaptive, iteratively functioning and support vector machine (SVM)–based ECG signal processing method. After a conventional pre-filtering step, the characteristic waves (QRS, P, T) from the ECG signal are localized. The implemented event estimation and recognition method uses an SVM-based unified neural network (UNN) in order to determine the most relevant parameters. A UNN-based preliminary ECG analyzer system has been created to reduce the searching space of the optimization algorithm. The obtained model parameters were determined by a relation between objective function minimization and robustness of the solution. The gained information allows an iterative filtering in permanent concordance with the aimed processing manner. Using these methods for one channel records from the MIT-BIH database, the detection rate of QRS complexes is above 99.9%. The negative influence of various noise types, like 50/60 Hz power line, abrupt baseline shift or drift, and low sampling rate was almost completely eliminated. The obtained wave locations and types can form a solid base to improve diagnosis performance in clinical environment. References [1] K. Minami, H. Nakajima, and T. Yoyoshima, “Real time discrimination of the ventricular tachyarrhythmia with Fourier-transform neural network”, IEEE Trans. Biomed. Eng., vol. 46, pp. 179–185, 1999. [2] M. Lagerholm, C. Peterson, G. Braccini, L. Edenbrandt and L. Sornmo, “Clustering ECG complexes using Hermite functions and self-organizing maps”, IEEE Trans. Biomed. Eng., vol. 47, pp. 838–847, 2000. [3] S. Osowski and L. T. Hoai, “ECG beat recognition using fuzzy hybrid neural network”, IEEE Trans. Biomed. Eng., vol. 48, pp. 1265–1271, 2001. [4] S. Osowski, L. T. Hoai and T. Markiewicz, “Support Vector Machine-Based Expert System for Reliable Heartbeat Recognition”, IEEE Trans. Biomed. Eng., vol. 51, pp. 582–589, 2004. [5] V. Vapnik, “Statistical Learning Theory”, New York: Wiley, 1998. [6] A. Smola and B. Scholkopf, “A tutorial on support vector regression”, Royal Holloway College, Univ. London, London, U.K., NeuroColt Tech. Rep. NV2-TR-1998-030, 1998. [7] Q. Xue, Y. H. Hu and W. J. Tompkins, “Neural-Network-Based Adaptive Matched Filtering for QRS Detection”, IEEE Trans. Biomed. Eng., vol. 39, pp. 317–329, 1992. [8] G. Nave and A. Cohen, “ECG Compression Using Long-Term Prediction”, IEEE Trans. Biomed. Eng., vol. 40, pp. 877–885, 1993. [9] S. M. Szilágyi, Z. Benyó and L. Dávid, “ECG Signal Compression and Noise Distortion Effect Analysis”, Proc. World Congr. Med. Phys. Biomed. Eng., Sydney, 4391.pdf , 2003. [10] K. Crammer and Y. Singer, “On the learnability and design of output codes for multi-class problems”, Proc. 13th Conf. Comput. Learning Theory, pp. 35–46, 2000. [11] S. M. Szilágyi, Z. Benyó and L. Dávid, “Iterative ECG Filtering for Better Malfunction Recognition and Diagnosis”, Proc. 5th IFAC Symp. Modell. Contr. Biomed. Syst., Melbourne, pp. 295–300, 2003. [12] C. Burges, “A tutorial on support vector machines for pattern recognition in Knowledge Discovery and Data Mining”, in U. Fayyad, Norwell, MA: Kluwer, 1–43, 2000. [13] C. Li, C. Zheng and C. Tai, “Detection of ECG Characteristic Points Using Wavelet Transforms”, IEEE Trans. Biomed. Eng., vol. 42, pp 21–27, 1995. [14] S. M. Szilagyi, L. Szilagyi and L. David, “ECG signal compression using adaptive prediction”, Proc. 19th Ann. Int. Conf. of IEEE EMBS, Chicago, pp. 101–104, 1997. [15] S. M. Szilagyi, L. Szilagyi and L. David, “Comparison between neural-network-based adaptive filtering and wavelet transform for ECG characteristic points detection”, Proc. 19th Ann. Int. Conf. of IEEE EMBS, Chicago, pp. 272–274, 1997. [16] S. M. Szilagyi and L. Szilagyi, “Wavelet Transform and Neural-Network-Based Adaptive Filtering for QRS Detection”, Proc. World Congr. Med. Phys. Biomed. Eng., Chicago, pp. 1267–1270, 2000. [17] S. M. Szilagyi, L. Szilagyi, A. Frigy and A. Incze, “Holter Telemetry in the Study of Heart Rate Variability”, Romanian Heart Journal, Vol. 2., Nr. 6, pp. 143, 1996. [18] S. M. Szilagyi, “The Limits of Heart-Model-Based Computerized ECG Diagnosis”, Proc. World Congr. Med. Phys. Biomed. Eng., Chicago, pp. 1913–1916, 2000. [19] S. M. Szilagyi, Z. Benyo and L. David, “Iterative ECG Signal Filtering for Better Malfunction Recognition and Diagnosis”, Proc. 5th IFAC Symp. Modell. Contr. Biomed. Syst., Melbourne, pp. 301–303, 2003. [20] S. M. Szilagyi, L. Szilagyi, A. Frigy, L. K. Gorog, S. E. Laszlo and Z. Benyo, “3D Heart Simulation and Recognition of Various Events”, Proc. 27th Ann. Int. Conf. of IEEE EMBS, Shanghai, pp. 4038–4041, 2005. [21] Z. Benyo and L. Czinege, “Computer Analysis of Dynamic Systems with Application in Physiology”, 15th World Congress of IMACS on Scientific Computation, Modelling and Applied Mathematics, Berlin, Vol. III., pp. 663–668, 1997. [22] S. M. Szilagyi, “Non-Linear Adaptive Prediction Based ECG Signal Filtering”, Proc. 21st Ann. Int. Conf. of IEEE EMBS, Atlanta, pp. 286, 1999. [23] L. Szilagyi, S. M. Szilagyi, A. Frigy, S. E. Laszlo, L. K. Gorog and Z. Benyo, “Quick QRS Complex Detection for On-Line ECG and Holter Systems”, Proc. 27th Ann. Int. Conf. of IEEE EMBS, Shanghai, pp. 3906–3908, 2005. [24] S. M. Szilagyi and L. Szilagyi, “Efficient ECG Signal Compression Using Adaptive Heart Model”, Proc. 21st Ann. Int. Conf. of IEEE EMBS, Istanbul, vol. 3, pp. 2125–2128, 2001. [25] S. M. Szilagyi, Z. Benyo and L. David, “ECG Signal Compression and Noise Distortion Effect Analysis”, Proc. World Congr. Med. Phys. Biomed. Eng., Sydney, 4391.pdf, 2003. |