Title: Epileptic seizure detection in EEG using improved entropy

Authors: Arumai Thangam Phareson Gini; Manuel Packiaselvam Flower Queen

Addresses: Department of EEE, Noorul Islam University, Kumaracoil, Thuckalay, India ' Department of Electrical and Electronics Engineering, Noorul Islam University, Kumara Coil, Thuckalay, India

Abstract: Epilepsy is a chronic disorder of the brain that impacts people all around the world. It is tremendously challenging to investigate the chronicled EEG signal and the analysis of epileptic activity is a time consuming procedure. In this article, we suggest a novel ANN based epileptic seizure detection with the help of the improved entropy feature. The anticipated technique includes stages like, pre-processing, feature abstraction and seizure detection. In the primary phase, we sample all the input information set. In second phase, a fuzzy entropy algorithm is utilised to abstract the features of EEG signal. Finally, we utilise artificial neural network for to recognise epilepsy seizures in exaggerated patient. Lastly, we associated the anticipated technique with prevailing technique for the perceiving epileptic sections. The function is utilised to compute the following parameters like accuracy, specificity, FAR, sensitivity, FRR, GAR which established the effectiveness of the anticipated epilepsy seizure recognition system.

Keywords: epileptic seizure; electroencephalogram; EEG; FAR; GAR; FRR; artificial neural networks; ANN.

DOI: 10.1504/IJBET.2020.108990

International Journal of Biomedical Engineering and Technology, 2020 Vol.33 No.4, pp.325 - 345

Received: 04 May 2017
Accepted: 04 Sep 2017

Published online: 07 Aug 2020 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article