Residual and bidirectional LSTM for epileptic seizure detection

Topics

ResBiLSTM · EEG classification · Seizure detection · one-dimensional residual neural network (1D ResNet) · Bidirectional LSTM (BiLSTM) · Hybrid deep learning model · Temporal modeling · Spatial feature extraction · University of Bonn EEG dataset · Temple University Hospital (TUH) EEG corpus · TUH seizure corpus (TUSZ) · Seizure detection · Seizure classification · Seizure classification · 10-fold cross-validation · fully connected layers · F1 score · residual learning for time-series signals