CLASSIFICATION OF EPILEPTIC AND HEALTHY INDIVIDUALS WITH RECURRENCE PARAMETER

Abstract

Epilepsy is a brain activity disorder that manifests itself with epileptic seizures. Although the reasons of epilepsy are not fully known, the diversity and variability of the individual make it difficult to diagnose epilepsy. For this, the diagnosis of epilepsy with computerized systems is one of the most popular research topics in recent years. Although many techniques and methods have been developed for this purpose, Electroencephalogram (EEG) signals are one of the most preferred and most basic ways to diagnose epileptic seizures or epilepsies because of their practicality and easy application. However, interpretation of EEG signals is not easy due to nonlinear and variable signal characteristics. This has led to the preference of nonlinear methods as well as traditional methods in the study of EEG signals. It can be seen from the literature that non-linear methods give very successful results in previous studies. In this study; EEG signals from healthy and epileptic subjects were represented by recurrence parameters obtained by recurrence plot. The features extracted from the recurrence plot are applied to multi-layered artificial neural networks, k-nearest neighbors, and support vector machines classifiers after feature selection process. Accordingly, the highest classification accuracy was achieved at around 97.05% when the multi-layered artificial neural network was used

Keywords

• recurrence parameters,recurrence quantification analysis,EEG,epilepsy,non-linear classification

References

1. H. Reynolds, “Epilepsy: The Disorder,” Epilepsy Atlas, pp. 15-27, 2005.
2. “Epilepsy Foundation”, received from the address: http://www.epilepsy.com/learn/epilepsy-101/what-epilepsy.
3. G. D. Cascino, “When drugs and surgery don't work”, Epilepsia, vol.9, pp. 79-84, 2008.
4. R. M. Rangayyan, Biomedical Signal Analysis, Canada: John Wiley & Sons, Inc., 2002.
5. R. G. Andrzejak, K. Lehnertz, F. Mormann, C. Rieke, P. David and C. E. Elger, “Indications of nonlinear deterministic and finite-dimensional structures in time series of brain electrical activity: Dependence on recording region and brain state”, Physical Review E, vol. 64, no.6, 2001.
6. T. Gautama, D. P. Mandic and M. M. V. Hulle., "Indications of nonlinear structures in brain electrical activity." Physical Review E, vol. 67, no. 4, 2003.
7. N. Kannathal, M. L. Choo, U. R. Acharya and P. K. Sadasivan, “Entropies for detection of epilepsy in EEG”, Computer methods and programs in biomedicine, vol. 80, no. 3, pp. 187-194, 2005.
8. T. Tzallas, M. G. Tsipouras and D. I. Fotiadis, “Automatic seizure detection based on time-frequency analysis and artificial neural networks”, Computational Intelligence and Neuroscience, 2007.

9. K. Polat and S. Güneş, “Classification of epileptiform EEG using a hybrid system based on decision tree classifier and fast Fourier transform”, Applied Mathematics and Computation, vol. 187, no. 2, pp. 1017-1026, 2007.
10. U. R. Acharya, S. V. Sree, S. Chattopadhyay, W. Yu and P. C. A. Ang, “Application of recurrence quantification analysis for the automated identification of epileptic EEG signals”, International journal of neural systems, vol. 21, no. 3, pp. 199-211, 2011.
11. J. P. Eckmann, S. O. Kamphorst and D. Ruelle, “Recurrence plots of dynamical systems”, Europhysics Letters, vol.4, no.9, 973, 1987.
12. H. Liu and R. Setiono, “Chi2: Feature selection and discretization of numeric attributes”, Tools with artificial intelligence, pp. 388-391, 1995.
13. E. Alpaydın, Yapay Öğrenme, Boğaziçi Üniversitesi Yayınevi, İstanbul, Mart 2011.
14. E. Öztemel, Yapay Sinir Ağları, Papatya Yayıncılık, İstanbul, 2006. [15] S. Abe, Support Vector Machines for Pattern Classification, Springer, Newyork, 2010.
15. C. Cortes and V. Vapnik, “Support Vector Networks”, Machine Learning, vol. 20, pp. 273-297, 1995.
16. D. T. Larose, Discovering Knowledge in Data, A John Wiley & Sons Inc. Publication, Newyork, 2005.
17. Kutlu and C. Kose, “Detection of epileptic seizure from EEG signals by using recurrence quantification analysis” presented at the 22nd IEEE Signal Processing and Communications Applications Conference (SIU), pp. 1387-1390, Trabzon, 2014.