Epileptic seizure detection in EEG using mutual information-based best individual feature selection

• Published in: Expert systems with applications Vol. 193; p. 116414
• Main Authors: Hassan, Kazi Mahmudul, Islam, Md. Rabiul, Nguyen, Thanh Thi, Molla, Md. Khademul Islam
• Format: Journal Article
• Language: English
• Published: New York Elsevier Ltd 01.05.2022; Elsevier BV
• Subjects: Algorithms; Convulsions & seizures; Electroencephalogram; Electroencephalography; Empirical mode decomposition; Epilepsy; Mathematical analysis; Multi-layer perceptron neural network; Multilayer perceptrons; Multilayers; Mutual information-based best individual feature; Neural networks; Neurological diseases; Seizure; Seizures; Multi-layer perceptron neural network; Empirical mode decomposition; Seizure; Mutual information-based best individual feature; Electroencephalogram; Epilepsy
• ISSN: 0957-4174; 1873-6793
• DOI: 10.1016/j.eswa.2021.116414

Epilepsy is a group of neurological disorders that affect normal brain activities and human behavior. Electroencephalogram based automatic epileptic seizure detection has significant applications in epilepsy treatment and medical diagnosis. In this study, a novel epileptic seizure detection method is proposed with a combination of empirical mode decomposition, mutual information-based best individual feature (MIBIF) selection algorithm and multi-layer perceptron neural network. Initially, fixed length EEG epochs are decomposed into amplitude and frequency-modulated components called intrinsic mode functions (IMFs). Three features named ellipse area of second-order difference plot, variance and fluctuation index are calculated from first few IMFs. The most significant features are then selected from the calculated features using the MIBIF algorithm to produce a final feature set. Later, the generated feature set is fed into the multi-layer perceptron neural network (MLPNN) classifier. Two well-known benchmark epileptic EEG datasets are used in this study for experimental evaluations. The result of proposed approach shows a significant performance improvement compared to the recent state-of-the-art methods. •Electroencephalogram are decomposed into narrow-band by Empirical Mode Decomposition.•Three discriminating features are extracted from each narrow-band signal.•Dominant features are selected using Feature selection algorithm.•Seizure detection are accomplished using neural network based classifier.