ABSTRACT
This paper presents the method for the epilepsy classification based on electroencephalogram (EEG) signals. This method uses the spectrogram image of EEG signals and convolutional neural networks (CNN).The spectrogram image is obtained by mapping the spectrogram value of the results of the short-time Fourier Transform (STFT) to the RGB color map. The best spectrogram for CNN is determined based on the length of fast Fourier transform (FFT) and window length in the windowing technique. The proposed method is evaluated with the epileptic EEG signal datasets. The experimental results show CNN works optimally with spectrogram images from STFT on the window length of 128 and the length of FFT of 128. In these conditions, the performance of CNN in the classification of epilepsy is better than others and able to compete with existing methods.
References
- Health Organization. 2019. Epilepsy. Available at: https://www.who.int/news-room/fact-sheets/detail/epilepsy. (Accessed: 8th December 2019)Google Scholar
- Scheffer, I. E. et al.2018. ILAE classification of the epilepsies: position paper of the ILAE Commission for Classification and Terminology. Zeitschrift fur Epileptologie, 31, 4, 296--306.Google ScholarCross Ref
- 'EEG time series: epileptic EEG Data'. EEG time series download page. Available at: http://epileptologie-bonn.de/cms/front_content. php?idcat=193&lang=3.(Accessed: 5th December 2019)Google Scholar
- Nicolaou, N. and Georgiou, J. 2012. Detection of epileptic electroencephalogram based on Permutation Entropy and Support Vector Machines. Expert Systems with Applications, 39, 1, 202--209. Google ScholarDigital Library
- Song, Y. and Zhang, J. 2013. Automatic recognition of epileptic EEG patterns via Extreme Learning Machine and multiresolution feature extraction. Expert Systems with Applications, 40, 14, 5477--5489.Google ScholarCross Ref
- Patidar, S. and Panigrahi, T. 2017. Detection of epileptic seizure using Kraskov entropy applied on tunable-Q wavelet transform of EEG signals. Biomedical Signal Processing and Control, 34, 1, 74--80.Google ScholarCross Ref
- Li, Z., Feng, L., Wang, Y., Wang, C. and Bai, H. 2017. Hardware design of multiclass SVM classification for epilepsy and epileptic seizure detection. IET Circuits, Devices & Systems, 12, 1, 108--115.Google Scholar
- Yoshimura, H., Shimizu, T., Hori, M., Iwai, Y. and Kishida, S. 2012. EEG Signals Classification by Using an Ensemble TPUnit Neural Networks for the Diagnosis of Epilepsy. International Journal of Machine Learning and Computing 2, 6, 758--761.Google ScholarCross Ref
- Orhan, U., Hekim, M. and Ozer, M. 2011. EEG signals classification using the K-means clustering and a multilayer perceptron neural network model. Expert Systems with Applications, 38, 10, 13475--13481. Google ScholarDigital Library
- Zhu, G., Li, Y., Wen, P., Wang, S. and Zhong, N. 2013. Unsupervised classification of epileptic EEG signals with multi scale K-means algorithm. Lecture Notes in Computer Science 8211 LNAI, 158--167. Google Scholar
- Samiee, K., Kovács, P. and Gabbouj, M. 2015. Epileptic seizure classification of EEG time-series using rational discrete short-time fourier transform. IEEE Transactions on Biomedical Engineering 62, 2, 541--552.Google ScholarCross Ref
- Alçin, Ö. F. et al. 2016. Multi-category EEG signal classification developing time-frequency texture features based Fisher Vector encoding method. Neurocomputing 218, 251--258. Google ScholarDigital Library
- Hassan, A. R. and Subasi, A. 2016. Automatic identification of epileptic seizures from EEG signals using linear programming boosting. Computer Methods and Programs in Biomedicine, 136, 65--77. Google ScholarDigital Library
- Djemili, R., Bourouba, H. and Amara Korba, M. C. 2016. Application of empirical mode decomposition and artificial neural network for the classification of normal and epileptic EEG signals. Biocybernetics and Biomedical Engineering, 36, 1, 285--291.Google ScholarCross Ref
- Li, Y. et al. 2018. Epileptic Seizure Classification of EEGs Using Time-Frequency Analysis Based Multiscale Radial Basis Functions. IEEE Journal of Biomedical and Health Informatics, 22, 2, 386--397.Google ScholarCross Ref
- Yildirim, O. et al. 2019. Automated detection of diabetic subject using pre-trained 2D-CNN models with frequency spectrum images extracted from heart rate signals. Computers in Biology and Medicine, 113, 103387.Google ScholarCross Ref
- Yuan, L. and Cao, J. 2018. Patients' EEG Data Analysis via Spectrogram Image. International Conference on Intelligent Decision Technologies 13--21. doi:10.1007/978-3-319-59421-7Google Scholar
- Alhadhrami, E., Al-Mufti, M., Taha, B. and Werghi, N. 2019. Learned Micro-Doppler Representations for Targets Classification Based on Spectrogram Images. IEEE Access, 7, 139377--139387.Google ScholarCross Ref
- Andrzejak, R. G. et al. 2001. 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 -Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics, 64, 6, 8Google ScholarCross Ref
- Prieto-Guerrero, A. and Espinosa-Paredes, G. 2019. Linear signal processing methods and decay ratio estimation. Linear and Non-Linear Stability Analysis in Boiling Water Reactors (Elsevier Ltd, 2019). doi:10.1016/b978-0-08-102445-4.00006-0Google ScholarCross Ref
- Kehtarnavaz, N. 2008. Frequency Domain Processing. Digital Signal Processing System Design, 1, 175--196.Google ScholarCross Ref
- Buck, O. A. V. R. W. S. and J. R. 1999. Discrete-Time Signal Processing - 2ed - Oppenheim.pdf. in 468--471 (Upper Saddle River, NJ: Prentice Hall, 1999). Google Scholar
- Alex Krizhevsky, Ilya Sutskever, and G. E. H. 2017. Convolutional Neural Networks ImageNet Classification with Deep. communication of the ACM, 60, no. JUNE, 84--90.Google Scholar
- Xavier Glorot. 2010. Understanding the difficulty of training deep feedforward neural networks. in International Conference on Artificial Intelligence and Statistics (AISTATS), 9, 249--256.Google Scholar
- He, K., Zhang, X., Ren, S. and Sun, J. 2015. Delving deep into rectifiers: Surpassing human-level performance on imagenet classification. Proceedings of the IEEE International Conference on Computer Vision 2015 Inter, 1026--1034. Google Scholar
- Dumoulin, V. and Visin, F. 2018. A guide to convolution arithmetic for deep learning. 1--31.Google Scholar
- Ioffe, S. and Szegedy, C. 2015. Batch normalization: Accelerating deep network training by reducing internal covariate shift.32nd International Conference on Machine Learning, ICML 2015, 1, 448--456. Google Scholar
- Nagi, J. et al. 2011. Max-pooling convolutional neural networks for vision-based hand gesture recognition. 2011 IEEE International Conference on Signal and Image Processing Applications, ICSIPA 2011 342--347. doi:10.1109/ICSIPA.2011.6144164Google Scholar
- Bishop, C. M. and Pattern. 2006. Pattern Recognition and Machine Learning. springer. Google Scholar
- Pe, A. and Lozano, J. A. 2010. Sensitivity Analysis of k - Fold Cross Validation in Prediction Error Estimation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 32, 3, 569--575. Google ScholarDigital Library
- Jiao, Y. and Du, P. 2016. Performance measures in evaluating machine learning based bioinformatics predictors for classifications. Quantitative Biology, 4, 4, 320--330.Google ScholarCross Ref
Index Terms
Epileptic EEG Signal Classification Using Convolutional Neural Networks Based on Optimum Window Length and FFT's Length
Comments