Abstract
As one of the important tools of epilepsy diagnosis, the electroencephalogram (EEG) is noninvasive and presents no traumatic injury to patients. It contains a lot of physiological and pathological information that is easy to obtain. The automatic classification of epileptic EEG is important in the diagnosis and therapeutic efficacy of epileptics. In this article, an explainable graph feature convolutional neural network named WTRPNet is proposed for epileptic EEG classification. Since WTRPNet is constructed by a recurrence plot in the wavelet domain, it can fully obtain the graph feature of the EEG signal, which is established by an explainable graph features extracted layer called WTRP block. The proposed method shows superior performance over state-of-the-art methods. Experimental results show that our algorithm has achieved an accuracy of 99.67% in classification of focal and nonfocal epileptic EEG, which proves the effectiveness of the classification and detection of epileptic EEG.
- [1] . 2011. Application of recurrence quantification analysis for the automated identification of epileptic EEG signals. International Journal of Neural Systems 21, 03 (June 2011), 199–211. https://doi.org/10.1142/S0129065711002808
PMID: 21656923 .Google ScholarCross Ref
- [2] . 2002. Comparison of wavelet transform and FFT methods in the analysis of EEG signals. Journal of Medical Systems 26, 3 (
June 2002), 241–247. https://doi.org/10.1023/A:1015075101937 Google ScholarDigital Library
- [3] . 2012. Nonrandomness, nonlinear dependence, and nonstationarity of electroencephalographic recordings from epilepsy patients. Physical Review E 86, 4 Pt 2 (Oct. 2012), 046206. https://doi.org/10.1103/PhysRevE.86.046206Google Scholar
Cross Ref
- [4] . 2020. Recognition of neurodegenerative diseases with gait patterns using double feature extraction methods. In Proceedings of the 2020 4th International Conference on Intelligent Computing and Control Systems (ICICCS’20). 332–338. https://doi.org/10.1109/ICICCS48265.2020.9120920Google Scholar
- [5] . 2017. Tunable-Q wavelet transform based multiscale entropy measure for automated classification of epileptic EEG signals. Applied Sciences 7, 4 (April 2017), 385. https://doi.org/10.3390/app7040385Google Scholar
Cross Ref
- [6] . 2016. A novel approach for automated detection of focal EEG signals using empirical wavelet transform. Neural Computing and Applications 29, 8 (Nov. 2016), 47–57. https://doi.org/10.1007/s00521-016-2646-4 Google Scholar
Digital Library
- [7] . 2017. Reduction of massive EEG datasets for epilepsy analysis using Artificial Neural Networks. In Proceedings of the 2017 10th International Conference on Human System Interactions (HSI’17). 137–143. https://doi.org/10.1109/HSI.2017.8005015Google Scholar
Cross Ref
- [8] . 2014. Recurrence plot analysis of HRV for exposure to low-frequency noise. In Frontiers of Energy, Materials and Information Engineering. Advanced Materials Research, Vol. 1044. Trans Tech Publications Ltd., 1251–1257. https://doi.org/10.4028/www.scientific.net/AMR.1044-1045.1251Google Scholar
- [9] . 2019. Fast density peak clustering for large scale data based on kNN. Knowledge-Based Systems 187 (Jan. 2019), 104824. https://doi.org/10.1016/j.knosys.2019.06.032Google Scholar
Cross Ref
- [10] . 2020. BLOCK-DBSCAN: Fast clustering for large scale data. Pattern Recognition 109, 1 (Jan. 2020), 107624. https://doi.org/10.1016/j.patcog.2020.107624Google Scholar
- [11] . 2019. KNN-BLOCK DBSCAN: Fast clustering for large-scale data. IEEE Transactions on Systems, Man, and Cybernetics: Systems 51, 6 (Dec. 2019), 3939–3953. https://doi.org/10.1109/TSMC.2019.2956527Google Scholar
- [12] . 1995. Recurrence Plots of Dynamical Systems.
World Scientific Series on Nonlinear Science Series A , Vol. 16. World Scientific, 441–445. https://doi.org/10.1142/9789812833709_0030Google Scholar - [13] . 2014. ILAE official report: A practical clinical definition of epilepsy. Epilepsia 55, 4 (April 2014), 475–482. https://doi.org/10.1111/epi.12550Google Scholar
Cross Ref
- [14] . 2012. Nonlinear analysis of EEG using fractal dimension and approximate entropy. In Materials Science and Information Technology II. Advanced Materials Research, Vol. 532. Trans Tech Publications Ltd., 988–992. https://doi.org/10.4028/www.scientific.net/AMR.532-533.988Google Scholar
- [15] . 1982. Automatic recognition of epileptic seizures in the EEG. Electroencephalography and Clinical Neurophysiology 54, 5 (Nov. 1982), 530–540. https://doi.org/10.1016/0013-4694(82)90038-4Google Scholar
Cross Ref
- [16] . 2012. Chaotic behavior of transistor circuits. In Applications of Chaos and Nonlinear Dynamics in Science and Engineering, Vol. 2. Springer, Berlin, Germany, 59–91. https://doi.org/10.1007/978-3-642-29329-0_4Google Scholar
- [17] . 2016. Epileptic seizure detection in EEG signals using tunable-Q factor wavelet transform and bootstrap aggregating. Computer Methods and Programs in Biomedicine 137 (Dec. 2016), 247–259. https://doi.org/10.1016/j.cmpb.2016.09.008Google Scholar
Cross Ref
- [18] . 2020. Localization of epileptic seizure focus by computerized analysis of fMRI recordings. Brain Informatics 7, 1 (Oct. 2020), Article 13, 13 pages. https://doi.org/10.1186/s40708-020-00114-0Google Scholar
Cross Ref
- [19] . 2020. Recurrence quantification analysis of electrooculography signal to a control question test: A new approach for the detection of deception. Biomedical Engineering: Applications, Basis and Communications 32, 04 (Aug. 2020), 2050029. https://doi.org/10.4015/S1016237220500295Google Scholar
Cross Ref
- [20] . 2005. EMD and instantaneous phase detection of structural damage. In Hilbert-Huang Transform and Its Applications.
Interdisciplinary Mathematical Sciences , Vol. 5. World Scientific, 227–262. https://doi.org/10.1142/9789812703347_0011Google Scholar - [21] . 2014. Analysis of spike waves in epilepsy using Hilbert-Huang transform. Journal of Medical Systems 39, 1 (Dec. 2014), 1–13. https://doi.org/10.1007/s10916-014-0170-6 Google Scholar
Digital Library
- [22] . 2019. Convolutional neural network and guided filtering for SAR image denoising. Remote Sensing 11, 6 (Feb. 2019), 702–720. https://doi.org/10.3390/rs11060702Google Scholar
Cross Ref
- [23] . 2017. SAR image denoising via sparse representation in Shearlet domain based on continuous cycle spinning. IEEE Transactions on Geoscience and Remote Sensing 55, 5 (Feb. 2017), 2985–2992. https://doi.org/10.1109/TGRS.2017.2657602Google Scholar
Cross Ref
- [24] . 2019. A deep residual neural network based framework for epileptogenesis detection in a rodent model with single-channel EEG recordings. In Proceedings of the 2019 12th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI’19). 1–6. https://doi.org/10.1109/CISP-BMEI48845.2019.8965693Google Scholar
Cross Ref
- [25] . 2020. Emotion classification in Parkinson’s disease EEG using RQA and ELM. In Proceedings of the 2020 16th IEEE International Colloquium on Signal Processing (CSPA’20). 290–295. https://doi.org/10.1109/CSPA48992.2020.9068709Google Scholar
- [26] . 2011. How to avoid potential pitfalls in recurrence plot based data analysis. International Journal of Bifurcation and Chaos 21, 04 (April 2011), 1003–1017. https://doi.org/10.1142/S0218127411029008Google Scholar
Cross Ref
- [27] . 2007. Recurrence plots for the analysis of complex systems. Physics Reports 438, 5 (Jan. 2007), 237–329. https://doi.org/10.1016/j.physrep.2006.11.001Google Scholar
Cross Ref
- [28] . 2020. EEG channel-selection method for epileptic-seizure classification based on multi-objective optimization. Frontiers in Neuroscience 14 (June 2020), 593. https://doi.org/10.3389/fnins.2020.00593Google Scholar
Cross Ref
- [29] . 2016. Evaluation of selected recurrence measures in discriminating pre-ictal and inter-ictal periods from epileptic EEG data. Physics Letters A 380, 16 (2016), 1419–1425. https://doi.org/10.1016/j.physleta.2016.02.024Google Scholar
Cross Ref
- [30] . 2011. EEG signals classification using the K-means clustering and a multilayer perceptron neural network model. Expert Systems with Applications 38, 10 (Sept. 2011), 13475–13481. https://doi.org/10.1016/j.eswa.2011.04.149 Google Scholar
Digital Library
- [31] . 2005. Automated prediction of epileptic seizures in rats with recurrence quantification analysis. In Proceedings of the 2005 Annual International Conference of the IEEE Engineering in Medicine and Biology Society.153–156. https://doi.org/10.1109/iembs.2005.1616365Google Scholar
- [32] . 2017. A new approach to characterize epileptic seizures using analytic time-frequency flexible wavelet transform and fractal dimension. Pattern Recognition Letters 94 (July 2017), 172–179. https://doi.org/10.1016/j.patrec.2017.03.023 Google Scholar
Digital Library
- [33] . 2016. DWT based detection of epileptic seizure from EEG signals using naive Bayes and k-NN classifiers. IEEE Access 4 (
June 2016), 7716–7727. https://doi.org/10.1109/ACCESS.2016.2585661Google ScholarCross Ref
- [34] . 2021. Automated seizure diagnosis system based on feature extraction and channel selection using EEG signals. Brain Informatics 8, 1 (Feb. 2021), 16. https://doi.org/10.1186/s40708-021-00123-7Google Scholar
- [35] . 2017. Classification of focal and non focal epileptic seizures using multi-features and SVM classifier. Journal of Medical Systems 41, 10 (Sept. 2017), 160. https://doi.org/10.1007/s10916-017-0800-x Google Scholar
Digital Library
- [36] . 2016. A novel robust diagnostic model to detect seizures in electroencephalography. Expert Systems with Applications 56 (Sept. 2016), 116–130. https://doi.org/10.1016/j.eswa.2016.02.040 Google Scholar
Digital Library
- [37] . 2018. Automatic seizure detection using three-dimensional CNN based on multi-channel EEG. BMC Medical Informatics and Decision Making 18, 5 (Dec. 2018), 111. https://doi.org/10.1186/s12911-018-0693-8Google Scholar
Cross Ref
- [38] . 2009. Biomedical data processing using HHT: A review. In New Advances in Biomedical Signal Processing. Springer, Berlin, Germany, 335–352. https://doi.org/10.1007/978-3-540-89506-0_16Google Scholar
- [39] . 2021. Epilepsy EEG classification based on convolution support vector machine. Journal of Medical Imaging and Health Informatics 11, 1 (Jan. 2021), 25–32. https://doi.org/10.1166/jmihi.2021.3259Google Scholar
Cross Ref
- [40] . 2017. AR based quadratic feature extraction in the VMD domain for the automated seizure detection of EEG using random forest classifier. Biomedical Signal Processing and Control 31 (Jan. 2017), 550–559. https://doi.org/10.1016/j.bspc.2016.10.001Google Scholar
Cross Ref
- [41] . 2020. Classification of epileptic IEEG signals by CNN and data augmentation. In Proceedings of the 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP’20). 926–930. https://doi.org/10.1109/ICASSP40776.2020.905Google Scholar
Cross Ref
- [42] . 2018. Epileptic seizure detection based on EEG signals and CNN. Frontiers in Neuroinformatics 12 (Dec. 2018), 95. https://doi.org/10.3389/fninf.2018.00095Google Scholar
Cross Ref
Index Terms
WTRPNet: An Explainable Graph Feature Convolutional Neural Network for Epileptic EEG Classification
Recommendations
Automated EEG analysis of epilepsy: A review
Epilepsy is an electrophysiological disorder of the brain, characterized by recurrent seizures. Electroencephalogram (EEG) is a test that measures and records the electrical activity of the brain, and is widely used in the detection and analysis of ...
Feature Extraction and Classification of Mental EEG for Motor Imagery
ICNC '09: Proceedings of the 2009 Fifth International Conference on Natural Computation - Volume 02Electroencephalography (EEG) recognition was one of the key technology in brain-computer interface (BCI). For motor imagery EEG, a new EEG recognition algorithm (DWT-BP algorithm) which combined discrete wavelet transform (DWT) with BP neural network ...
Classification of epileptic EEG recordings using signal transforms and convolutional neural networks
AbstractThis paper describes the analysis of a deep neural network for the classification of epileptic EEG signals. The deep learning architecture is made up of two convolutional layers for feature extraction and three fully-connected layers ...
Highlights- This paper analyzes the use of a deep neural network for epileptic EEG signal classification.






Comments