Epileptic seizure detection in EEG signals using deep learning: LSTM and bidirectional LSTM

Area of Science:

• Neuroscience and Biomedical Engineering
• Artificial Intelligence in Healthcare

Background:

• Epileptic seizures are a neurological disorder characterized by abnormal brain activity.
• Accurate detection of epileptic seizures from electroencephalogram (EEG) signals is crucial for diagnosis and treatment.
• Existing methods for EEG-based seizure detection often face challenges in accuracy and efficiency.

Purpose of the Study:

• To develop a novel, automatic method for detecting epileptic seizures in EEG signals.
• To enhance the accuracy and performance of seizure detection using advanced deep learning techniques.
• To validate the effectiveness of a hybrid Long Short-Term Memory (LSTM) and bidirectional LSTM (BiLSTM) network for epileptic EEG signal classification.

Main Methods:

• EEG signals were decomposed into different sub-bands using discrete wavelet transform (DWT).
• A hybrid deep learning model combining one layer of Long Short-Term Memory (LSTM) and one layer of bidirectional LSTM (BiLSTM) was developed.
• The proposed LSTM-BiLSTM network was trained and evaluated for the classification of epileptic EEG signals.

Main Results:

• The developed automatic method achieved a high accuracy of 100% in detecting epileptic seizures.
• The hybrid LSTM-BiLSTM network demonstrated superior performance compared to other existing methods.
• Experimental results confirmed the effectiveness of the proposed model in classifying epileptic EEG signals.

Conclusions:

• The novel DWT and hybrid LSTM-BiLSTM deep learning approach provides an effective and highly accurate method for automatic epileptic seizure detection.
• This hybrid model represents a significant advancement in the field of automated EEG analysis for epilepsy.
• The findings suggest promising potential for clinical application in real-time seizure monitoring and diagnosis.