Optimized Seizure Detection Algorithm: A Fast Approach for Onset of Epileptic in EEG Signals Using GT Discriminant Analysis and K-NN Classifier

Background: Epilepsy is a severe disorder of the central nervous system that predisposes the person to recurrent seizures. Fifty million people worldwide suffer from epilepsy; after Alzheimer’s and stroke, it is the third widespread nervous disorder.Objective: In this paper, an algorithm to detect the onset of epileptic seizures based on the analysis of brain electrical signals (EEG) has been proposed. ۸۴۴ hours of EEG were recorded form ۲۳ pediatric patients consecutively with ۱۶۳ occurrences of seizures. Signals had been collected from Children’s Hospital Boston with a sampling frequency of ۲۵۶ Hz through ۱۸ channels in order to assess epilepsy surgery. By selecting effective features from seizure and non-seizure signals of each individual and putting them into two categories, the proposed algorithm detects the onset of seizures quickly and with high sensitivity.Method: In this algorithm, L-sec epochs of signals are displayed in form of a thirdorder tensor in spatial, spectral and temporal spaces by applying wavelet transform. Then, after applying general tensor discriminant analysis (GTDA) on tensors and calculating mapping matrix, feature vectors are extracted. GTDA increases the sensitivity of the algorithm by storing data without deleting them. Finally, K-Nearest neighbors (KNN) is used to classify the selected features.Results: The results of simulating algorithm on algorithm standard dataset shows that the algorithm is capable of detecting ۹۸ percent of seizures with an average delay of ۴.۷ seconds and the average error rate detection of three errors in ۲۴ hours.Conclusion: Today, the lack of an automated system to detect or predict the seizure onset is strongly felt.

Keywords:

EEG Signals , Epileptic Seizure , General Tensor Discriminant Analysis (GTDA) , K-NN , Wavelet Transform