When a brain produces signals which are larger than it would normally produce, causes an epileptic seizure. This epileptic seizure caused by neurological disorder can be detected by analysis Electroencephalogram (EEG) signal using an efficient algorithm. In this paper, a new EEG signal analysis method for identifying epileptic seizure is proposed. The suggested method uses Variational Mode Decomposition (VMD), Neighborhood Component Feature Selection (NCFS), and machine learning classification approach. The signal processing will start with extracting 44 features from each EEG signal using VMD technique, and NCFS method has been employed to select best features from the extracted features. The obtained simulation results prove that the proposed EEG signal analysis algorithm is able to detect the epileptic seizure achieving the best accuracy.
Volume 11 | 06-Special Issue
Pages: 1401-1411