Background: Automatic Recognition of Epileptic Seizure from EEG Signal based on Discrete Wavelet Transform. In traditional methods such as visual scanning, epilepsy detection is a time-consuming process which requires high accuracy and experience for analyzing the whole length of data. The main objective of this study is automatic diagnosis of epileptic seizures via discrete wavelet transform.
Methods and findings: The data sets used in this work are electroencephalogram (EEG) signals. The classification of EEG signals proposed in this paper for diagnosis of epileptic seizures is based on wavelet transform. The method is composed of three steps: a) wavelet transform based feature extraction, b) feature space dimension reduction based on scatter matrices and c) classification. The proposed approach, applied on EEG data sets, belong to three categories: a) healthy persons, b) persons with epilepsy during a seizure-free period and c) persons with epilepsy during a seizure.
Conclusion: The classification in this work is more accurate than previous studies. Conducted simulation shows the effectiveness of proposed method.
Arefe Moharrami and Abdolaziz Kalte