癲癇影響全世界約0.8%的人口，而在台灣就有大約十五萬人患有癲癇，雖然某部分的癲癇患者嚴重癲癇發作的頻率不高，但在生活中發生意外的可能以及旁人的想法皆會讓癲癇患者產生不小的心理壓力。過去大多數針對癲癇的研究僅止於癲癇檢測，但癲癇檢測對於癲癇患者在日常生活中的幫助極其有限，無法有效的讓癲癇患者避開危險。雖然使用腦電圖預測癲癇的研究在近年來有成熟的發展，然而腦電圖不論是在穿戴上以及外觀皆不利於在日常配戴的條件。因此本研究提出使用心電訊號搭配深度學習網路來預測癲癇發作，並針對臨床的穿戴式用途做調整。相較於過去使用心律變異度參數來預測癲癇，本研究所使用的深度學習網路在即時性、泛化性上更加具有優勢，以時間卷積網路搭配自我注意力機制預測癲癇有效的標記出發作前的訊號，同時也能偵測到癲癇發作的時間。本研究僅使用網路開源資料PhysioNet.com的七筆癲癇資料，需適當的處理資料以及調整訓練參數來彌補少量訓練中的缺點。最終在七筆資料中有四筆成功預測出顯著的發作前訊號，可以提前10到15分鐘預測到癲癇的發作，證實了深度學習方法應用於癲癇預測任務上，即使使用少量資料仍有不錯的效果，且深度學習方法還有助於在臨床應用上的即時預測效果。

Epilepsy affects approximately 0.8% of the global population. In Taiwan, more than 150,000 people suffer from epilepsy. Although most patients with epilepsy (PWE) have slight and infrequent seizures, the possibility of accidents and other people’s thoughts cause psychological pressure on PWE. Most studies on epilepsy have focused on seizure detection, but seizure detection is extremely limited in helping PWE in their daily lives. Electroencephalography (EEG) has been used to predict seizures in recent years. However, EEG has disadvantages in daily use in terms of wearing comfort and appearance. Therefore, this study proposes a methodology that uses an electrocardiogram (ECG) to predict seizures and adjust it for wearable use. Compared with the use of heart rhythm variability (HRV) parameters to predict seizures in the past, the deep learning network used in this study has more advantages in real-time and generalization. The temporal convolutional networks and self-attention mechanism are used to predict seizures effectively, and the preictal signal can also be detected to influence the use of seizure prediction. In this study, seven epilepsy data from PhysioNet.com, an open-source web site, were used, and the data had to be processed properly, and the training parameters had to be adjusted to compensate for the shortcomings of few-shot learning. In the end, four of the seven data successfully predicted preictal signals significantly, and the seizures could be predicted up to 15 min in advance, which proves that the deep learning method could be applied for seizure prediction tasks even in the case of training with a small amount of data.

[1] R. S. Fisher, C. Acevedo, A. Arzimanoglou, et al., “ILAE Official Report: A practical clinical definition of epilepsy,” Epilepsia, vol. 55, no. 4, pp. 475-482, 2014.
[2] “ポケモン騒動を検証する,” Fab. 19, 2004. Accessed on: Jul. 23, 2020. [Online]. Available: https://web.archive.org/web/20080113092133/http://home-aki.cool.ne.jp/soudou.htm
[3] S. Shorvon, “The etiologic classification of epilepsy,” Epilepsia, vol. 52, no. 6, pp. 1052-1057, 2011.
[4] A. M. Pack, “Epilepsy Overview and Revised Classification of Seizures and Epilepsies,” CONTINUUM: Lifelong Learning in Neurology, vol. 25, no. 2, pp. 306–321, 2019.
[5] National Institute of Neurological Disorders and Stroke and National Institutes of Health, “What Are the Different Kinds of Seizures?” in The Epilepsies and Seizures: Hope Through Research, 2004, ch. 4, pp. 10–11. [Online]. Available: https://www.ninds.nih.gov/Disorders/Patient-Caregiver-Education/Hope-Through-Research/Epilepsies-and-Seizures-Hope-Through
[6] J. Extercatte, G. de Haan and A. Gaitatzis, “Teaching Video NeuroImages: Frontal opercular seizures with Jacksonian march,” Neurology, vol. 84, no. 11, pp. e83-e84, 2015.
[7] R. Mameniškienė and P. Wolf,“ Epilepsia partialis continua: A review,” Seizure, vol. 44, pp. 74–80, 2017.
[8] W. H. Theodore, R. J. Porter, P. Albert, K. Kelley, E. Bromfield, O. Devinsky, and S. Sato, “The secondarily generalized tonic‐clonic seizure: a videotape analysis,” Neurology, vol. 44, no. 8, pp. 1403-1403, 1994.
[9] M. M. Zack and R. Kobau, “National and State Estimates of the Numbers of Adults and Children with Active Epilepsy — United States, 2015,” MMWR. Morbidity and Mortality Weekly Report, vol. 66, no. 31, pp. 821–825, 2017.
[10] E. Trinka, P. Kwan, B. Lee and A. Dash, “Epilepsy in Asia: Disease burden, management barriers, and challenges,” Epilepsia, vol. 60, pp. 7-21, 2018.
[11] Shuo-Bin Jou, “認識癲癇”, Mackay Memorial Hospital, Health Education Information, Accessed on: Jun. 16, 2020. [Online]. Available: https://www.mmh.org.tw/taitam/neuro/index4_7.html
[12] Statistics Department of Ministry of Health and Welfare of Taiwan. Available: https://dep.mohw.gov.tw/DOS/mp-113.html
[13] World Health Organization, “Global status report on alcohol and health,” 2011. Available: https://www.who.int/substance_abuse/publications/global_alcohol_report/msbgsruprofiles
[14] M. Ruíz-García et al., “Clinical guideline: pre-operative evaluation of epilepsy surgery,” Revista Mexicana de Neurociencia, vol. 20, no. 2, 2019.
[15] M. Mathews, M. Linskey and D. Binder, “William P. van Wagenen and the first corpus callosotomies for epilepsy,” Journal of Neurosurgery, vol. 108, no. 3, pp. 608-613, 2008.
[16] Souza, Karla Teixeira, Guerreiro, M. Marilisa, and Guerreiro, Carlos A.M., “Accident risk in patients with epilepsy,” Arquivos de Neuro-Psiquiatria, 2008.
[17] F. van den Bongard, H. Hamer, R. Sassen and C. Reinsberger, “Sport and physical activity in epilepsy,” Deutsches Aerzteblatt Online, 2020.
[18] K. Gordon, J. Dooley and P. Brna, “Epilepsy and activity-A population-based study,” Epilepsia, vol. 51, no. 11, pp. 2254-2259, 2010.
[19] T. Tomson, L. Nashef and P. Ryvlin, “Sudden unexpected death in epilepsy: current knowledge and future directions,” The Lancet Neurology, vol. 7, no. 11, pp. 1021-1031, 2008.
[20] S. D. Lhatoo, M. Nei, M. Raghavan, M. Sperling, B. Zonjy, N. Lacuey, and O. Devinsky, “Nonseizure SUDEP: Sudden unexpected death in epilepsy without preceding epileptic seizures,” Epilepsia, vol. 57, no. 7, pp. 1161–1168, 2016.
[21] M. Poh, T. Loddenkemper, C. Reinsberger, N. C. Swenson, S. Goyal, J. R. Madsen, and R. W. Picard, “Autonomic changes with seizures correlate with postictal EEG suppression,” Neurology, vol. 78, no. 23, pp. 1868–1876, 2012.
[22] T. Tomson, R. Surges, R. Delamont, S. Haywood, and D. C. Hesdorffer, “Who to target in sudden unexpected death in epilepsy prevention and how? Risk factors, biomarkers, and intervention study designs,” Epilepsia, vol. 57, pp. 4–16, 2016.
[23] M. J. Maguire, C. F. Jackson, A. G. Marson, and S. J. Nevitt, “Treatments for the prevention of Sudden Unexpected Death in Epilepsy (SUDEP),” Cochrane Database of Systematic Reviews, 2016.
[24] H. Blumenfeld, “Impaired consciousness in epilepsy,” The Lancet Neurology, vol. 11, no. 9, pp. 814-826, 2012.
[25] D. Chen, S. Wan, and F. S. Bao, “Epileptic Focus Localization Using Discrete Wavelet Transform Based on Interictal Intracranial EEG,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 25, no. 5, pp. 413–425, 2017.
[26] M. Sharma, R. B. Pachori, U. R. Acharya et al., “A new approach to characterize epileptic seizures using analytic time-frequency flexible wavelet transform and fractal dimension,” Pattern Recognition Letters, vol. 94, pp. 172-179, 2017.
[27] N. Arunkumar, K. Ramkumar, V. Venkatraman et al., “Classification of Focal and Non Focal EEG Using Entropies,” Pattern Recognition Letters, vol. 94, pp. 112-117, 2017.
[28] Z. K. Gao, Q. Cai, Y. Z. Yang et al., “Visibility Graph from Adaptive Optimal Kernel Time-Frequency Representation for Classification of Epileptiform EEG,” International Journal of Neural Systems, vol. 27, no. 4, 2017.
[29] U. R. Acharya, S. L. Oh, Y. Hagiwara et al., “Deep convolutional neural network for the automated detection and diagnosis of seizure using EEG signals,” Computers in Biology and Medicine, vol. 100, pp. 270-278, 2018.
[30] E. Alickovic, J. Kevric, and A. Subasi, “Performance evaluation of empirical mode decomposition, discrete wavelet transform, and wavelet packed decomposition for automated epileptic seizure detection and prediction,” Biomedical Signal Processing and Control, vol. 39, pp. 94–102, 2018.
[31] P. Rajna, B. Clemens, E. Csibri, E. Dobos, A. Geregely, M. Gottschal, I. György, Á. Horváth, F. Horváth, L. Mezöfi, I. Velkey, J. Veres, and E. Wagner, “Hungarian multicentre epidemiologic study of the warning and initial symptoms (prodrome, aura) of epileptic seizures,” Seizure, vol. 6, no. 5, pp. 361–368, 1997.
[32] Web of Science, database: SCI-E, SSCI, A&HCI, ESCI, key word: (epilepsy | seizure) & prediction; (epilepsy | seizure) & detection. Available: http://apps.webofknowledge.com/
[33] F. Mormann, T. Kreuz, C. Rieke et al., “On the predictablility of epileptic seizures,” Clinical Neurophysiology, vol. 116, no. 3, pp. 569-587, 2005.
[34] C. Meisel and K. Bailey, “Identifying signal-dependent information about the preictal state: A comparison across ECoG, EEG and EKG using deep learning,” EBioMedicine, vol. 45, pp. 422-431, 2019.
[35] E. Alickovic, J. Kevric, and A. Subasi, “Performance evaluation of empirical mode decomposition, discrete wavelet transform, and wavelet packed decomposition for automated epileptic seizure detection and prediction,” Biomedical Signal Processing and Control, vol. 39, pp. 94–102, 2018.
[36] P. J. Karoly, D. R. Freestone, R. Boston, D. B. Grayden, D. Himes, K. Leyde, U. Seneviratne, S. Berkovic, T. O’Brien, and M. J. Cook, “Interictal spikes and epileptic seizures: their relationship and underlying rhythmicity,” Brain, vol. 139, no. 4, pp. 1066–1078, 2016.
[37] B. H. Brinkmann, J. Wagenaar, D. Abbot, P. Adkins, S. C. Bosshard, M. Chen, Q. M. Tieng, J. He, F. J. Muñoz-Almaraz, P. Botella-Rocamora, J. Pardo, F. Zamora-Martinez, M. Hills, W. Wu, I. Korshunova, W. Cukierski, C. Vite, E. E. Patterson, B. Litt, and G. A. Worrell, “Crowdsourcing reproducible seizure forecasting in human and canine epilepsy,” Brain, vol. 139, no. 6, pp. 1713–1722, 2016.
[38] Κ. Μ. Tsiouris, V. C. Pezoulas, M. Zervakis, S. Konitsiotis, D. D. Koutsouris, and D. I. Fotiadis, “A Long Short-Term Memory deep learning network for the prediction of epileptic seizures using EEG signals,” Computers in Biology and Medicine, vol. 99, pp. 24–37, 2018.
[39] N. D. Truong, A. D. Nguyen, L. Kuhlmann, M. R. Bonyadi, J. Yang, S. Ippolito, and O. Kavehei, “Convolutional neural networks for seizure prediction using intracranial and scalp electroencephalogram,” Neural Networks, vol. 105, pp. 104–111, 2018.
[40] L. Kuhlmann, K. Lehnertz, M. P. Richardson, B. Schelter, and H. P. Zaveri, “Seizure prediction — ready for a new era,” Nature Reviews Neurology, vol. 14, no. 10, pp. 618–630, 2018.
[41] K. Gadhoumi, J. Lina, F. Mormann and J. Gotman, “Seizure prediction for therapeutic devices: A review,” Journal of Neuroscience Methods, vol. 260, pp. 270-282, 2016.
[42] K. Myers, S. Sivathamboo and P. Perucca, “Heart rate variability measurement in epilepsy: How can we move from research to clinical practice?” Epilepsia, vol. 59, no. 12, pp. 2169-2178, 2018.
[43] K. Vandecasteele, T. De Cooman, Y. Gu et al., “Automated Epileptic Seizure Detection Based on Wearable ECG and PPG in a Hospital Environment,” Sensors, vol. 17, no. 10, 2017.
[44] M. Poh, “Continuous assessment of epileptic seizures with wrist-worn biosensors,” Ph.D. dissertation, Philosophy in Electrical and Medical Engineering, Massachusetts Institute of Technology, 2011.
[45] A. Ulate-Campos, F. Coughlin, M. Gaínza-Lein, I. S. Fernández, P. Pearl, and T. Loddenkemper, “Automated seizure detection systems and their effectiveness for each type of seizure,” Seizure, vol. 40, pp. 88–101, 2016.
[46] K. Fujiwara, M. Miyajima, T. Yamakawa et al., “Epileptic Seizure Prediction Based on Multivariate Statistical Process Control of Heart Rate Variability Features,” IEEE Transactions on Biomedical Engineering, vol. 63, no. 6, pp. 1321-1332, 2016.
[47] M. K. Moridani and H. Farhadi, “Heart rate variability as a biomarker for epilepsy seizure prediction,” Bratislava Medical Journal, vol. 118, no. 01, pp. 3–8, 2017.
[48] Pavei, Jonatas, Heinzen, “Early Seizure Detection Based on Cardiac Autonomic Regulation Dynamics,” Frontiers Physiology, vol. 8, No. 765, 2017.
[49] Billeci, Lucia, Marino, “Patient-specific seizure prediction based on heart rate variability and recurrence quantification analysis,” Plos One, vol. 13, no. 9, 2018.
[50] Meisel, Christian, Bailey, “Identifying signal-dependent information about the preictal state: A comparison across ECoG, EEG and EKG using deep learning,” Ebiomedicine, vol. 45, pp. 422-431, 2019.
[51] M. Ribeiro, A. E. Lazzaretti, and H. S. Lopes, “A study of deep convolutional auto-encoders for anomaly detection in videos,” Pattern Recognition Letters, vol. 105, pp. 13-22, 2018.
[52] Goldberger AL, Amaral LAN, Glass L et al., “PhysioNet: Components of a New Research Resource for Complex Physiologic Signals,” Circulation, vol. 101, no. 23, pp. 215-220, 2000.
[53] Al-Aweel, I. C., Krishnamurthy, K. B., Hausdorff et al., “Postictal heart rate oscillations in partial epilepsy,” Neurology, vol. 53, no. 7, pp. 1590-1590, 1999.
[54] J. Pan and W. Tompkins, “A Real-Time QRS Detection Algorithm,” IEEE Transactions on Biomedical Engineering, vol. -32, no. 3, pp. 230-236, 1985.
[55] H. Sedghamiz, “Complete Pan-Tompkins Implementation ECG QRS Detector,” Matlab Central: Community Profile, 2014.
[56] C. Ufongene, R. El Atrache, T. Loddenkemper, and C. Meisel, “Electrocardiographic changes associated with epilepsy beyond heart rate and their utilization in future seizure detection and forecasting methods,” Clinical Neurophysiology, vol. 131, no. 4, pp. 866-879, 2020.
[57] J. Devlin, M. W. Chang, K. Lee, and K. Toutanova, “Bert: Pre-training of deep bidirectional transformers for language understanding,” arXiv preprint arXiv:1810.04805, 2018.
[58] M. Jordan, “Serial Order: A Parallel Distributed Processing Approach,” San Diego, Calif University, 1986.
[59] J. Elman, “Finding Structure in Time,” Cognitive Science, vol. 14, no. 2, pp. 179-211, 1990.
[60] S. Hochreiter and J. Schmidhuber, “Long short-term memory,” Neural computation, vol.9, no. 8, pp. 1735-1780, 1997.
[61] J. Chung, C. Gulcehre, K. Cho, and Y. Bengio, “Empirical evaluation of gated recurrent neural networks on sequence modeling,” arXiv preprint arXiv:1412.3555, 2014.
[62] C. Lea, M. D. Flynn, R. Vidal, A. Reiter, and G. D. Hager, “Temporal Convolutional Networks for Action Segmentation and Detection,” 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017.
[63] S. Bai, J. Kolter, and V. Koltun, “An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling,” arXiv preprint arXiv:1803.01271, 2018.
[64] A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez et al. “Attention is all you need,” arXiv preprint arXiv:1706.03762, 2017
[65] N. S. Keskar, D. Mudigere, J. Nocedal, M. Smelyanskiy, and P. T. P. Tang, “On large-batch training for deep learning: Generalization gap and sharp minima,” arXiv preprint arXiv:1609.04836, 2016.