Real-Time Clinical EEG-Based Epileptic Seizure Detection Using a Hybrid CNN–LSTM–SVM Deep Learning Model
Journal of Intelligent Decision Making and Information Science · 09.09.2026 · 3 (18s) · 2411-2423
DOI: 10.17762/jidmis.v3.5992 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Intelligent Decision Making and Information Science, online erschienen am 09.09.2026. Verfasst von Aakanksha Joshi. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Electroencephalogram (EEG) data are complicated and non-stationary; accurately and promptly detecting epileptic seizures continues to be a major problem in neurological healthcare. The necessity for sophisticated automated seizure detection systems is highlighted by the labour-intensive, time-consuming, and subjectively variable nature of conventional manual EEG interpretation. Using EEG recordings obtained from a government hospital in accordance with recognised clinical guidelines, this study suggests a hybrid deep learning framework for real-time epileptic seizure identification. After preprocessing the acquired EEG signals to eliminate noise and artefacts, they were segmented into fixed-length windows. Continuous Wavelet Transform (CWT) was used to create time-frequency scalogram images in order to efficiently capture the temporal and spectral features of the EEG data. A hybrid CNN–LSTM–SVM architecture was then used to process these scalograms. The Convolutional Neural Network (CNN) automatically extracted discriminative spatial features, the Long Short-Term Memory (LSTM) network learned temporal dependencies among sequential features, and the Support Vector Machine (SVM) carried out the final seizure classification. To provide a reliable performance evaluation, the suggested framework was trained and assessed utilising distinct training and testing datasets. Standard performance criteria, including accuracy, precision, recall, F1-score, confusion matrix, and training-validation loss analysis, were used to evaluate the experiment. The acquired results show that the suggested hybrid framework reliably detects seizures with enhanced classification performance and successfully learns complicated EEG patterns. An intelligent decision-support system for real-time epilepsy monitoring and clinical diagnosis is developed through the integration of scalogram-based feature representation, hybrid deep learning, and real-time clinical EEG data.
Zitieren
APA
Joshi, A. (2026). Real-Time Clinical EEG-Based Epileptic Seizure Detection Using a Hybrid CNN–LSTM–SVM Deep Learning Model. Journal of Intelligent Decision Making and Information Science, 3(18s), 2411-2423. https://doi.org/10.17762/jidmis.v3.5992
Vancouver
Joshi A. Real-Time Clinical EEG-Based Epileptic Seizure Detection Using a Hybrid CNN–LSTM–SVM Deep Learning Model. Journal of Intelligent Decision Making and Information Science. 2026;3(18s):2411-2423. doi:10.17762/jidmis.v3.5992
BibTeX
@article{Joshi2026,
author = {Joshi, Aakanksha},
title = {Real-Time Clinical EEG-Based Epileptic Seizure Detection Using a Hybrid CNN–LSTM–SVM Deep Learning Model},
journal = {Journal of Intelligent Decision Making and Information Science},
year = {2026},
volume = {3},
number = {18s},
pages = {2411--2423},
doi = {10.17762/jidmis.v3.5992}
}
EUVIMED