Deep Learning Models in the Early Prediction of Epileptic Seizures
Alkadhim Journal for Computer Science · 25.03.2026 · 22
DOI: 10.61710/kjcs.v4i1.149 ↗Kurz gefasst
Fachartikel in der Zeitschrift Alkadhim Journal for Computer Science, online erschienen am 25.03.2026. Verfasst von Shahad Sallah. Die Arbeit nennt 20 Literaturangaben.
Abstract (Originalsprache)
Seizure prediction remains an open problem in the field of clinical neuroscience, given its unpredictable nature and the risk profile associated with it. The present study undertakes a comparative assessment of the performance of various architectures of deep learning, including Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM), Graph Neural Network (GNN), and Transformer, in predicting epileptic seizures 15 minutes in advance using EEG signals. A standard protocol has been adopted for the experiments, including preprocessing techniques such as filtering, normalization, and segmentation, in addition to feature extraction and balancing, using the CHB-MIT dataset. The performance of the models has been measured using metrics such as accuracy, specificity, reproducibility, F1-score, and area under the receiver operating characteristic curve (AUC).
Zitieren
APA
Sallah, S. (2026). Deep Learning Models in the Early Prediction of Epileptic Seizures. Alkadhim Journal for Computer Science, 22. https://doi.org/10.61710/kjcs.v4i1.149
Vancouver
Sallah S. Deep Learning Models in the Early Prediction of Epileptic Seizures. Alkadhim Journal for Computer Science. 2026:22. doi:10.61710/kjcs.v4i1.149
BibTeX
@article{Sallah2026,
author = {Sallah, Shahad},
title = {Deep Learning Models in the Early Prediction of Epileptic Seizures},
journal = {Alkadhim Journal for Computer Science},
year = {2026},
pages = {22},
doi = {10.61710/kjcs.v4i1.149}
}
EUVIMED