A scheme combining feature fusion and hybrid deep learning models for epileptic seizure detection and prediction

Topics

Seizure detection · Seizure prediction · Feature extraction · CNN-GRU-AM model · Convolutional neural network · Gated recurrent unit · Attention mechanism · Discrete Wavelet Transform · EEG signal analysis · Time-frequency feature extraction · Nonlinear features · EEG signal analysis · CHB-MIT dataset · 10-fold cross-validation · Hybrid deep learning model · Sensitivity and specificity evaluation · Area under the receiver operating characteristic curve (AUC) · Feature extraction · Seizure prediction · Patient monitoring system