Epileptic seizure detection based on The Limited Penetrable visibility graph algorithm and graph properties

Abstract

Introduction: Epileptic seizure detection is a key step for both researchers and epilepsy specialists for epilepsy assessment due to the non-stationariness and chaos in the electroencephalogram (EEG) signals. Current research is directed toward the development of an efficient method for epilepsy or seizure detection based the limited penetrable visibility graph (LPVG) algorithm. Materials and Methods: Single channel EEG signals are mapped into the LPVGs and then 14 features are calculated from these graphs. Then some features are selected using Sequential forward feature selection method and given to error-correcting output codes (ECOC) to classify signals into three groups of healthy, seizure free (interictal) and during a seizure (ictal) groups. Results: Experimental results show our method can classify normal, ictal and interictal groups with a high accuracy of 96.31%. Conclusion: The proposed method is fast and easy. Comparison the performance of the proposed method with other automatic seizure detection method also shows our method has better performance.