Deep Learning based Reliable Early Epileptic Seizure Predictor

Seizure prediction has a great influence on epileptic patients' quality of life. In this paper, a novel deep learning based patient-specific epileptic seizure predictor using electroencephalogram (EEG) is proposed. The presented method is based on detecting the preictal state and differentiating it from the predominant interictal state. EEG features extraction and classification are achieved in a single automated system that is able to extract the most discriminative features from the raw EEG. The proposed approach gains the benefit of the convolutional neural network in extracting the spatial features from the multichannel EEG signals and the bidirectional recurrent neural network in predicting the occurrence of the incoming seizure earlier than the current methods. High accuracy of 99.6%, low false alarm rate of 0.004 h-1and early seizure prediction of one hour make the proposed method the most efficient among the state of the art. Robustness of the system is verified using an effective test method.

[1]  Liang-Gee Chen,et al.  Seizure prediction based on classification of EEG synchronization patterns with on-line retraining and post-processing scheme , 2011, 2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[2]  Jeffrey M. Hausdorff,et al.  Physionet: Components of a New Research Resource for Complex Physiologic Signals". Circu-lation Vol , 2000 .

[3]  Reza Tafreshi,et al.  Predicting Epileptic Seizures in Scalp EEG Based on a Variational Bayesian Gaussian Mixture Model of Zero-Crossing Intervals , 2013, IEEE Transactions on Biomedical Engineering.

[4]  Jürgen Schmidhuber,et al.  Framewise phoneme classification with bidirectional LSTM and other neural network architectures , 2005, Neural Networks.

[5]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[6]  Mojtaba Bandarabadi,et al.  Epileptic seizure prediction using relative spectral power features , 2015, Clinical Neurophysiology.

[7]  Mohamad Sawan,et al.  Towards accurate prediction of epileptic seizures: A review , 2017, Biomed. Signal Process. Control..

[8]  Michalis E. Zervakis,et al.  Discrimination of Preictal and Interictal Brain States from Long-Term EEG Data , 2017, 2017 IEEE 30th International Symposium on Computer-Based Medical Systems (CBMS).

[9]  Mahmoud I. Khalil,et al.  Epileptic seizure prediction using zero-crossings analysis of EEG wavelet detail coefficients , 2016, 2016 IEEE Conference on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB).

[10]  Haidar Khan,et al.  Focal Onset Seizure Prediction Using Convolutional Networks , 2018, IEEE Transactions on Biomedical Engineering.

[11]  J. H. Cross,et al.  ILAE Official Report: A practical clinical definition of epilepsy , 2014, Epilepsia.

[12]  Ali H. Shoeb,et al.  Application of machine learning to epileptic seizure onset detection and treatment , 2009 .

[13]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.