Abstract
Automatic detection and classification of arrhythmias based on ECG signals are important to cardiac-disease diagnostics. The ability of the ECG classifier to identify arrhythmias accurately is based on the development of robust techniques for both feature extraction and classification. A classifier is developed based on using wavelet, transforms for extracting features and then using a radial basis function neural network (RBFNN) to classify the arrhythmia. Six energy descriptors are derived from the wavelent coefficients, over a single-beat interval from the ECG signal. Nine different continuous and discrete wavelet transforms, are considered for obtaining the feature vector. An RBFNN adapted to detect and classify life-threatening arrhythmias is then used to classify the feature vector. Classification results are based on 159 arrhythmia, files obtained from three different sources. Classification results indicate the potential for wavelet based energy descriptors to distinguish the main features of the signal and thereby enhance the classification scheme. The RBFNN classifier appears to be well suited to classifying the arrhythmia, owing to the feature vectors' linear inseparability, and tendency to cluster. Utilising the Daubechies wavelet transform, an overall correct classification of 97.5% is obtained, with 100% correct classification for both ventricular fibrillation and ventricular tachycardia.
Similar content being viewed by others
References
Barro, S., Ruiz, Cabello, R., andMira, J. (1989): ‘Algorithmic sequential decision-making in the frequency domain for life threatening ventricular arrhythmias and imitative artifacts: a diagnostic system’,J. Biomed. Eng.,11, pp. 320–328
Burrus, C., Gopinath, A., andGuo, H. (1998): ‘Introduction to wavelets and wavelet transforms: a primer (Prentice Hall)
Cabello, D., Barro, S., Salceda, J. M., Ruiz, R., andMira, J. (1991): ‘Fuzzy K-nearst neighbor classifiers for ventricular arrhythmia detection’,Int. J., Biomed. Comput.,27, pp. 77–93
Charllis, R. E., andKittney, R. I. (1990): ‘Biomedical signal processing (in four parts). Part 1 time domian methods’,Med. Biog. Eng. & Comput.,28, pp. 509–524
Chen, S. W., Clarkson, P. M., andFan, Q. (1996): ‘A robust sequential detection algorithm for cardiac arrhythmia classification’,IEEE Trans.,BME-43, pp. 1120–1125
Clayton, R. Murray, A., andCampbell, R. (1991): ‘Changes in the surface ECG frequency spectrum during the onset of ventricular fibrilation’. Proc. Computer Cardiology 1986, IEEE Computer Society Press, pp. 515–518
Clayton, R. H. Murray, A., andCampbell, R. W. F. (1993): ‘Comparisons of four techniques for recognition of ventricular fibrillation from the surface ECG’,Med. Biol. Eng. Comput.,31, pp. 111–117
Fukunga, K. (1990): ‘Statistical pattern recognition’, 2nd ed. (Academic Press)
Grimm, W., Flores, B. F., andMarchlinski, F. E. (1992): ‘Electrocardiographically documented unnecessary, spontaneous shocks in 241 patients with implantable cardioverter defibrillators’,Pacing Clin. Electrophysiol.,15, pp. 1667–1673
Gritzali, F., Frangakis, G., andPapakonstantinou, G. (1989): ‘Detection of the P and T waves in an ECG’,Comput. Biomed. Res. 22, pp. 83–91
Gurunathan, S., Hsia, P. W., Lawton J., andHua, D. P. (1998): ‘Vector magnitude using orthogonal ECG leads during ventricular fibrillation is associated with defibrillation outcome’,Biomed. Instrum. Technol.,32, pp. 48–61.
Haykin, S. (1994): ‘Neural networks a comprehensive foundation’, (IEEE Press)
Howitt, I. (1995): ‘Radial basis function methodology for use in digital communications’. PhD dissertation, University of California Davis
Khadra, L., Al-Fahoum, A. S., andAl-Nashash, H. (1997): ‘Detection of life-threatening cardiac arrhythmias using the wavelet transformation’,Med. Biol. Eng. Comput. 35, pp 626–632
Mallat, S. (1991): ‘Zero-crossings of a wavelet transform’,IEEE Trans. Info. Theory,37, pp. 1019–1033
Meij, S. andZeelenberg, C. (1987): ‘A fast real time algorithm for the detection of ventricular fibrillation’. Proc. Computer Cardiology 1986, (IEEE Computer Society Press, Washington DC), pp. 707–710
Nogami, A., Takahashi, A., Naito, S., Tsuchio, Y., Oshima, S., Taniguchi, K., Nitta, J., Aonuma, K., andIesaka, Y. (1996): ‘Shortcut link between the fast and slow pathways and the mechanism of cure in atrioventricular nodal reentrant tachycardia by catheter ablation’,Pacing Clin. Electrophysiol. 19, pp. 1972–1977
Nygards, M., andHulting, J. (1977): ‘Recognition of ventricular fibrillation from power spectrum of ECG’, Proc. Computer Cardiology 1977, (IEEE Computer Society Press, Long Beach, California), pp. 393–397
Orlov, Y. S., Brodsky, M. A., Orlov, M. V., Allen, B. J., andWinters, R. J. (1995): ‘Is the time domain signal averaged electrocardiogram helpful in patients with ventricular tachcardia without apparent structural heart disease?’,Clin. Cardiol.,18, pp. 568–572
Pan, J., andTompkins, W. J. (1985): ‘A real-time QRS detection algorithm’,IEEE Trans.,BME-32, pp. 230–236
Popescu, M., Laskaris, N., Chiladakis, I., Stathopoulos, C., Cristea, P., Manolis, A., andBezerianos, A. (1998): ‘Beat-to-beat wavelet variance of the QRS complex as a marker of arrhythmogenic substrate in ventricular tachycardia patients’,Physiol. Meas.,19, pp. 77–92
Strang, G., andNguyen, T. (1996): ‘Wavelets and filter banks’ (Wellesley Cambridge Press)
Thakor, S., Webster, J., andTompkins W. J. (1984): ‘Estimation of QRS complex power spectra for design of a QRS filter’,IEEE Trans.,BME-31
Thakor, N. V., Zhu, Y. S., andPan, K. Y. (1990): ‘Ventricular tachycardia and fibrillation detection by a sequential hypothesis testing algorithm’,IEEE Trans.,BME-37, pp. 837–843
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Al-Fahoum, A.S., Howitt, I. Combined wavelet transformation and radial basis neural networks for classifying life-threatening cardiac arrhythmias. Med. Biol. Eng. Comput. 37, 566–573 (1999). https://doi.org/10.1007/BF02513350
Received:
Accepted:
Issue Date:
DOI: https://doi.org/10.1007/BF02513350