Deep Learning-Based Epilepsy Detection Using EEG Signals and One-Dimensional CNNs

Epilepsy is a neurological disorder characterized by recurrent seizures resulting from abnormal brain activity. Electroencephalography (EEG) is the gold standard for epilepsy diagnosis, but manual interpretation is time-consuming and subject to human error. This study presents a deep learning-based approach for automated epilepsy classification using EEG signals from the publicly available Bonn dataset. EEG segments were preprocessed using the Fast Fourier Transform (FFT) to extract frequency-domain features, including power spectral density, band-specific energy, spectral entropy, and dominant frequencies. A one-dimensional convolutional neural network (CNN) was developed in PyTorch to classify EEG signals into four categories: healthy (eyes open and closed), interictal, and ictal. The model architecture included convolutional layers, max-pooling, batch normalization, dropout, and fully connected layers. The network was trained using the Adam optimizer with early stopping, learning rate scheduling, and L۲ regularization to prevent overfitting. The model achieved a high validation accuracy of ۹۸.۷۵%, with consistent results across loss, accuracy, and mean squared error metrics. These results demonstrate the effectiveness of CNNs combined with FFT-based preprocessing in accurately detecting epileptic activity from EEG signals, supporting their potential for real-time and reliable clinical applications.

کلیدواژه ها:

Epilepsy ، EEG ، Convolutional Neural Network ، Deep Learning