The Application of Neural Networks in the Classification of the Electrocardiogram

  • C. D. Nugent
  • J. A. Lopez
  • N. D. Black
  • J. A. C. Webb
Part of the Studies in Fuzziness and Soft Computing book series (STUDFUZZ, volume 96)


The introduction of computers to assist with classification of the Electrocardiogram (ECG) is considered to be one of the earliest instances of computers in medicine. Over the past 4 decades, since its inception, research techniques in the given field have proliferated. Approaches adopted have included the use of rule based approaches such as Decision Trees, Fuzzy Logic and Expert Systems, to the use of Multivariate Statistical Analysis. The past decade has seen the approach of Neural Networks (NNs) being not only employed in the field, but a very popular and successful approach in comparison with the aforementioned techniques.


Feature Vector Feature Selection Feature Extraction Hide Layer Ventricular Repolarization 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    Tortora, J.G. and Grabowski, S.R. (1996), Principles of Anatomy and Physiology, 8th Edition, Harper Collins College Publishers, New York.Google Scholar
  2. [2]
    Berbari, E.J. (1995), “Principles of electrocardiography,” in: Bronzino, J.D. (Ed.), The Biomedical Engineering Handbook, CRC Press, Florida, pp. 181–190.Google Scholar
  3. [3]
    Wellens, H.J.J. (1990), “Electrocardiography past, present and future,” Annals of the New York Academy of Sciences, vol. 601 Electrocardiography, pp. 305–311.Google Scholar
  4. [4]
    Willems, J.L. (1990), “Quantitative electrocardiography standardization and performance evaluation,” Annals of the New York Academy of Sciences, vol. 601 Electrocardiography, pp. 329–342.Google Scholar
  5. [5]
    Blake, T.M. (1994), The Practice of Electrocardiography, 5th Edition, Humana Press, Totowa, New Jersey.CrossRefGoogle Scholar
  6. [6]
    Kors, J.A. and van Bemmel, J.H. (1990), “Classification methods for computerized interpretation of the electrocardiogram,” Meth. Inform. Med., vol. 29, pp. 330–336.PubMedGoogle Scholar
  7. [7]
    Van Bemmel, J.H., Zywietz, C., and Kors, J.A. (1990), “Signal analysis for ECG interpretation,” Meth. Inform. Med., vol. 29, pp. 317–329.PubMedGoogle Scholar
  8. [8]
    Stallman, F.W. and Pipberger, H.V. (1961), “Automatic recognition of electrocardiographic waves by digital computer,” Circ. Res., vol. 9, pp. 1138–1143.Google Scholar
  9. [9]
    Macfarlane, P. (1992), “Recent developments in computer analysis of ECGs,” Clinical Physiology, vol. 12, pp. 313–317.PubMedCrossRefGoogle Scholar
  10. [10]
    Willems, J.L., Abreu-Lima, C., Arnaud, P., van Bemmel, J.H., Brohet, C., Degani, R., Denis, B., Graham, I., van Herpen, G., Macfarlane, P.W., Michaelis, J., Moulopoulos, S., Poppl, S., and Zywietz, C. (1988), “Effect of combining electrocardiographic interpretation results on diagnostic accuracy,” European Heart Journal, vol. 9, pp. 1348–1355.PubMedGoogle Scholar
  11. [11]
    Jenkins, J (1981), “Computerized electrocardiography,” CRC Crit. Rev. Bioeng., vol. 6, pp. 307–350.Google Scholar
  12. [12]
    Rowlandson, I. (1990), “Computerized electrocardiography a historical perspective,” Annals of the New York Academy of Sciences, vol. 601 Electrocardiography, pp. 343–352.Google Scholar
  13. [13]
    Willems, J.L., Arnaud, P., van Bemmel, J.H., Degani, R., MacFarlane, P.W., and Zywietz, C. (1992), “Comparison of diagnostic results of ECG computer programs and cardiologists,” Computers in Cardiology, pp. 93–96.Google Scholar
  14. [14]
    Macfarlane, P. (1990), “A brief history of computer assisted electrocardiography,” Meth. Inform. Med., vol. 29, pp. 272–281.PubMedGoogle Scholar
  15. [15]
    Yang, T.F., Devine, B., and Macfarlane, P.W. (1993), “Deterministic logic versus software-based artificial neural networks in the diagnosis of atrial fibrillation,” J. Electrocardiology, vol. 26, pp. 90–94.Google Scholar
  16. [16]
    Bortolan, G. and Willems, J.L. (1993), “Diagnostic ECG classification based on neural networks,” Journal of Electrocardiology, vol. 26, pp. 75–79.PubMedGoogle Scholar
  17. [17]
    Bortolan, G., Brohet, C., and Fusaro, S. (1996), “Possibilities of using neural networks for ECG classification,” Journal of Electrocardiology, vol. 29, pp. 10–16.PubMedCrossRefGoogle Scholar
  18. [18]
    Waller, A.D. (1887), “A demonstration on man of electromotive changes accompanying the heart’s beat,” Journal Physiol., vol. 8, pp. 229–234.Google Scholar
  19. [19]
    Mathewson, F.A.L. and Jackh, H. (1955), “The Telecardiogram,” Amer. Heart Jnl., vol. 49, pp. 77–82. (Translation of: Einthoven, W. (1906) “Le Telecardiogramme,” Arch. Int. Physiol, vol. 4, pp. 132–161.)Google Scholar
  20. [20]
    Wilson, F.N., Macleod, A.G., and Barker, P.S. (1931), “The potential variations produced by the heart beat at the apices of einthoven’s triangle,” Amer. Heart Jnl., vol. 7, pp. 207–211.CrossRefGoogle Scholar
  21. [21]
    Goldberger, A.L. and Goldberger, E. (1994), Clinical electrocardiography a simplified approach, 5th Edition, Mosby, Missouri.Google Scholar
  22. [22]
    Report of Committee of the American Heart Association, The standardization of electrocardiographic nomenclature,” (1943), J.A.M.A, vol. 121, pp. 1347–1351.CrossRefGoogle Scholar
  23. [23]
    Nugent, C.D., Webb, J.A.C., Wright, G.T.H., and Black, N.D. (1998), “Electrocardiogram 1: pre-processing prior to classification,” Automedica, vol. 16, pp. 263–282.Google Scholar
  24. [24]
    Report of Committee on Electrocardiography, American Heart Association, Recommendations for standarization of electrcardiographic and vectorcardiographic leads,” (1954) Circulation, vol. 10, pp. 564–573.CrossRefGoogle Scholar
  25. [25]
    A report for health professionals by an ad hoc writing group of the Committee on Electrocardiography and Cardiac Electrophysiology of the Council on Clinical Cardiology, American Heart Association,” (1990), Circulation, vol. 81, pp. 730–739.CrossRefGoogle Scholar
  26. [26]
    Chandrasekaran, B. and Jain, A.K. (1974), “Quantization complexity and independent measurements,” IEEE Trans. Comp., pp. 102106.Google Scholar
  27. [27]
    Mucciardi, A.N. and Gose, E.E. (1971), “A comparison of seven techniques for choosing subsets of pattern recognition properties,” IEEE Trans. Comp., vol. C-20, no. 9, pp. 1023–1031.Google Scholar
  28. [28]
    Jain, U., Rautaharju, P.M., and Warren, J. (1981), “Selection of optimal features for classification of electrocardiograms,” J. Electrocardiography, vol. 14, no. 3, pp. 239–248.CrossRefGoogle Scholar
  29. [29]
    Cornfield, J., Dunn, R.A., Batchlor, C.D., and Pipberger, H.V. (1973), “Multigroup diagnosis of electrocardiograms,” Computers and Biomedical Research, vol. 6, pp. 97–120.PubMedCrossRefGoogle Scholar
  30. [30]
    Siddiqui, K.J., Greco, E.C., Kadri, N., Mohiuddin, S., and Sketch, M. (1993), “Best feature selection using successive elimination of poor performers,” Proc. Ann. Conf. IEEE Eng. in Med. and Biol., vol. 15, pp. 725–726.CrossRefGoogle Scholar
  31. [31]
    Jain, U. and Rautaharju, P.M. (1980), “Diagnostic accuracy of the conventional 12-lead and the orthogonal frank-lead electrocardiograms in detection of myocardial infarctions with classifiers using continuous and Bernoulli features,” J. Electrocardiology, vol. 13, no. 2, pp. 159–166.CrossRefGoogle Scholar
  32. [32]
    Chein, Y.T. and Fu, K.S. (1968), “Selection and ordering of feature observations in a pattern recognition system,” Information and Control, vol. 12, pp. 395–414.Google Scholar
  33. [33]
    Fukunaga, K. and Koontz, W.L.G. (1970), “Application of the Karhunen-Loeve expansion to feature selection and ordering,” IEEE Trans. Comp., vol. C-19, no. 4, pp. 311–318.Google Scholar
  34. [34]
    Udupa, J.K. and Murthy, I.S.N. (1980), “Syntactic approach to ECG rhythm analysis,” IEEE Trans. Biomed. Eng., vol. BME-27, no. 7, pp. 370–375.Google Scholar
  35. [35]
    Pietka, E. (1991), “Feature extraction in computerized approach to the ECG analysis,” Pattern Recognition, vol. 24, pp. 139–146.CrossRefGoogle Scholar
  36. [36]
    Pavlidis, T. (1973), “Waveform segmentation through functional approximation,” IEEE Trans. Comp., vol. C-22, no. 7, pp. 689697.Google Scholar
  37. [37]
    Willems, J.L. (1977), “Introduction to multivariate and conventional computer ECG analysis: pro’s and contra’s,” in: van Bemmel, J.H. and Willems, J.L. (Ed.), Trends in Computer-Processed Electrocardiograms, North-Holland Publishing Company, Holland, pp. 213–220.Google Scholar
  38. [38]
    Okajima, M., Okamoto, N., Yokoi, M., Iwatsuka, T., and Ohsawa, N. (1990), “Methodology of ECG interpretation in the Nagoya program,” Meth. Inform. Med., vol. 29, pp. 341–345.PubMedGoogle Scholar
  39. [39]
    Georgeson, S. and Warner, H. (1992), “Expert system diagnosis of wide complex tachycardia,” Computers in Cardiology, pp. 671674.Google Scholar
  40. [40]
    Held, C. and Kurien, J. (1992), “FLAPECAN: a knowledge based system in electrocardiogram diagnosis,” Proc. 5` h Annual Symposium on Computer-Based Medical Systems, pp. 648–655.Google Scholar
  41. [41]
    Degani, R. and Bortolan, G. (1987), “Fuzzy numbers in computerized electrocardiography,” Fuzzy Sets and Systems, vol. 24, pp. 345–362.CrossRefGoogle Scholar
  42. [42]
    Degani, R. (1992), “Computerized electrocardiogram diagnosis: fuzzy approach,” Methods of Information in Medicine, vol. 31, pp. 225–233.PubMedGoogle Scholar
  43. [43]
    Pipberger, H.V. et al. (1975), “Clinical application of a second generation electrocardiographic computer program,” Amer. J. Cardiol., vol. 35, pp. 597–608.PubMedCrossRefGoogle Scholar
  44. [44]
    Nugent, C.D., Webb, J.A.C., and Black, N.D. (2000), “Feature and classifier fusion for 12-lead ECG classification,” Medical Informatics & the Internet in Medicine, vol. 25, no. 3, pp. 225–235.CrossRefGoogle Scholar
  45. [45]
    Attikiouzel, Y. and deSilva, C.J.S. (1995), “Applications of neural networks in medicine,” Australasian Physical and Engineering Sciences in Medicine, vol. 18, pp. 158–164.PubMedGoogle Scholar
  46. [46]
    Lim, C.P., Harrison, R.F., and Kennedy, R.L. (1997), “Applications of autonomous neural network systems to medical pattern classification tasks,” Artificial Intelligence in Medicine, vol. 11, pp. 215–239.PubMedCrossRefGoogle Scholar
  47. [47]
    Dorffner, G. and Porenta, G. (1994), “On using feedforward neural networks for clinical diagnostic tasks,” Artificial Intelligence in Medicine, vol. 6, pp. 417–435.PubMedCrossRefGoogle Scholar
  48. [48]
    Bounds, D.G. and Lloyd, P.J. (1990), “A comparison of neural network and other pattern recognition approaches to the diagnosis of low back disorders,” Neural Networks, vol. 3, pp. 583–591.CrossRefGoogle Scholar
  49. [49]
    Lippmann, R. (1987), “An introduction to computing with neural nets,” IEEE ASAP Magazine, pp. 4–22.Google Scholar
  50. [50]
    Zahner, D.A. and Micheli-Tzanakou, E. (1995), “Artificial neural networks: Definitions, methods, applications,” in: Bronzino, J.D. (Ed.), The Biomedical Engineering Handbook, CRC Press, pp. 2699–2715.Google Scholar
  51. [51]
    Kartalopoulos, S.V. (1996), Understanding Neural Networks and Fuzzy Logic: Basic Concepts and Applications, IEEE Press, New York.Google Scholar
  52. [52]
    Haykin, S. (1994), Neural Networks: a Comprehensive Foundation, Macmillan College Publishing Company, New York.Google Scholar
  53. [53]
    Bortolan, G., Degani, R., and Willems, J.L. (1990), “Design of neural networks for classification of electrocardiographic signals,” Proc. Ann. IEEE Eng. in Med. and Bio. Soc., vol. 12, no. 3, pp. 1467–1468.Google Scholar
  54. [54]
    Willems, J.L., Arnaud, P., van Bemmel, J.H., Degani, R., MacFarlane, P.W., and Zwietz, C. (1990), “Common standards for quantitative electrocardiography: goals and main results,” Meth. Inform. Med., vol. 29, pp. 263–271.PubMedGoogle Scholar
  55. [55]
    Bortolan, G. and Willems, J.L. (1993), “Diagnostic ECG classification based on neural networks,” Journal of Electrocardiology, vol. 26, pp. 75–79.PubMedGoogle Scholar
  56. [56]
    Baxt, W.G. (1991), “Use of an artificial neural network for the diagnosis of myocardial infarction,” Annals of Internal Medicine, vol. 115, pp. 843–848.PubMedCrossRefGoogle Scholar
  57. [57]
    Harrison, R.F., Marshall, S.J., and Kennedy, R.L. (1991), “The early diagnosis of heart attacks: a neurocomputational approach,” Proc. Int. Joint Conf on Neural Networks, vol. 1, pp. 1–5.Google Scholar
  58. [58]
    Edenbrandt, L., Devine, B., and MacFarlane, P.W. (1992), “Neural networks for classification of ECG st-t segments,” J. Electrocardiology, vol. 25, pp. 167–173.CrossRefGoogle Scholar
  59. [59]
    Marques de Sa, J.P., Goncalves, A.P., Ferreira, F.O., and AbreuLima, C. (1994), “Comparison of artificial neural network based ECG classifiers using different features types,” Computers in Cardiology, pp. 545–547.Google Scholar
  60. [60]
    Xue, Q., Hu, Y.H., and Tompkins, W.J. (1990), “Training of ECG signals in neural network pattern recognition,” Proc. Ann. IEEE Eng. in Med. and Bio. Soc., pp. 1465–1466, vol. 12, no. 3.Google Scholar
  61. [61]
    Taur, J.S. and Kung, S.Y. (1993), “Prediction-based networks with ECG application,” Proc. IEEE Int. Conf. Neural Networks, vol. 3, pp. 1920–1925.CrossRefGoogle Scholar
  62. [62]
    Edenbrandt, L., Heden, B., and Pahlm, O. (1993), “Neural networks for analysis of ECG complexes,” J. Electrocardiology, vol. 26, pp. 74.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • C. D. Nugent
  • J. A. Lopez
  • N. D. Black
  • J. A. C. Webb

There are no affiliations available

Personalised recommendations