Epileptic Seizure Detection Based on Time Domain Features and Weighted Complex Network
Epileptic seizure detection is one of the important steps in diagnosis of epilepsy. Excellent automatic detection algorithm of epilepsy will help healthcare workers to better treat epilepsy patients, which has important study significance. In this paper, we proposed a new epileptic seizure detection method based on time domain features and weighted complex network of electroencephalogram (EEG) signals. Firstly, each EEG segment is divided into four sub-segments and each sub-segment is divided into thirty-two clusters. A set of time domain features is extracted from each cluster. Then, each set of this features is used as a node of complex network. Features sets are converted into weighted horizontal visibility graph. Thirdly, average weighted degree of complex network is extracted as the classification feature. Finally, average weighted degree is inputted into a linear classifier to classify epileptic EEG signals. The experimental result shows that the classification accuracy is up to 96.5%. The obtained result indicates that the proposed method is effective in epileptic seizure detection.
KeywordsEpileptic EEG signals Time domain features Weighted complex network Averaged weighted degree
This work was supported by the National Natural Science Foundation of China (Grant No. 61671220, 61701192, 61201428), the National Key Research and Development Program of China (No. 2016YFC0106000), the Project of Shandong Province Higher Educational Science and Technology Program, China (Grant No. J16LN07), the Shandong Province Key Research and Development Program, China (Grant No. 2016GGX101022), the Natural Science Foundation of Shandong Province, China, (Grant No. ZR2017QF004).
- 4.Kaplan, A.Y.: Segmental structure of EEG more likely reveals the dynamic multistability of the brain tissue than the continual plasticity one. In: Conference on Neural Information Processing, vol. 2, pp. 633–638. IEEE (1999)Google Scholar
- 15.Luque, B., Lacasa, L., Ballesteros, F., Liuque, J.: Horizontal visibility graphs: exact results for random time series. Phys. Rev. E 80 (2009). Article ID 046103Google Scholar
- 20.Meng, Q.F., Chen, S.S., Chen, Y.H., et al.: Automatic detection of epileptic EEG based on recurrence quantification analysis and SVM. Acta Physica Sinica 63(5) (2014). https://doi.org/10.7498/aps.63.050506
- 21.Wang, F., Meng, Q., Chen, Y., et al.: Feature extraction method for epileptic seizure detection based on cluster coefficient distribution of complex network. WSEAS Trans. Comput. 13, 351–360 (2014)Google Scholar