Epileptic Seizure Prediction and Source Localization using Entropy Analysis of Scalp EEG
Keywords:EEG, Epileptic Seizure, Pre-ictal period, Shannon wavelet entropy.
Epilepsy is one of most common neurological disorders that affects people of all ages and can cause unpredictable seizures which may even cause death. The prediction of epileptic activities thus can have a great impact in avoiding fatal injuries through early preparation with medicines and also in improving the efficacy of medicines. A technique for early prediction of epileptic seizure from EEG signal is proposed in this paper. The pre-ictal period of epileptic seizure clearly depicts a start of seizure and comparing the changes in entropy of EEG signals in different brain regions during the pre-seizure period, the proposed technique could successfully predict the seizure using entropy analysis. Moreover, the region of the epileptic activities was also localized by dividing the total brain into four topographic regions and by calculating the entropy from this four zones separately. Thus, this proposed technique has the potential to help the clinical neurologists to investigate seizure detection and treat the patient in a better way with less supervision and better accuracy.
Bangladesh Journal of Medical Physics Vol.11 No.1 2018 P 16-25