Classification of Epileptic Seizure EEG Based on Fully Connected Neural Network
Seok-Woo Jang · International Journal of Emerging Trends in Engineering Research · 2020
This paper proposes a method to classify epileptic seizure EEG and normal EEG using Fully Connected Neural Network (FCNN).Wavelet transform (WT) was used in the first step to extract features to be used in FCNN.In the second step, 24 features were extracted from the wavelet coefficients generated in the first step using a frequency distribution and a frequency variation.FCNN showed a classification performance of 99.99% when classifying epileptic seizure EEG and normal EEG using these 24 features.