EEG-Based Epileptic Seizure Detection Using a Hybrid CNN-LSTM Deep Learning Framework

Abstract

Epilepsy is a neurological disorder that affects millions of people worldwide. It causes repeated seizures due to abnormal electrical activity in the brain. Early and accurate seizure detection is important for proper diagnosis and treatment. Electroencephalography (EEG) is a non-invasive technique which is used to record brain signals and widely used for seizure analysis. However, manual analysis of EEG signals is difficult and time-consuming. In this study, a hybrid CNN-LSTM deep learning model is proposed for automatic epileptic seizure detection using EEG signals. The Epileptic Seizure Recognition dataset containing 11,500 samples was used for training and testing the model. Before training, the EEG data were normalized and prepared for deep learning analysis. The CNN layers were used to extract the important features from EEG signals, while the LSTM layer was used to learn patterns over time. The proposed model achieved an ac-curacy of 98.43% with high precision, recall, F1-score, and AUC values. The results show that the CNN-LSTM model could accurately classify seizure and non-seizure EEG signals. This study demonstrates that deep learning can be an effective tool for automatic seizure detection and can support future real-time healthcare and monitoring systems.