Deep Temporal Seizure Prediction A CNN-LSTM Hybrid for Enhanced EEG Signal Analysis
- DOI
- 10.2991/978-94-6463-718-2_37How to use a DOI?
- Keywords
- seizure prediction; EEG analysis; CNN-LSTM hybrid; deep learning; explainable AI; temporal feature extraction
- Abstract
The prediction of epilepsy, a significant field of investigation in neuroscience, has continuously focused on different machine learning models that depend on the evolving comprehension of EEG signal analysis. This paper introduces a hybrid deep learning framework that combines convolutional neural networks (CNNs) and long short-term memory (LSTMs) networks for dynamic seizure prediction. Lastly, our new hybrid network using CNN to extract features and use LSTM networks’ ability to find longer temporal dependencies. In order to mitigate inter-patient variability and reduce dependence on a single dataset, transfer learning and domain adaptation methods are applied to the model to encourage generalization across diverse datasets. Dense, strong against noise and artifacts, light attention mechanisms and quick preprocessing pipelines open space for real-time applications on low-power devices. Explainable ai modules were integrated into the model, allowing for a degree of clinical interpretability of the results, culminating in the one of the goals which was to help trust and deployment in the clinical setting. Consequently, through cross-validation, and time-frequency feature amalgamation, contributes toward the efficacy, scalability, and performance of the model. The hybrid approach achieves a significant improvement to EEG-based seizure prediction and ultimately will enable reliable and deployable clinical therapies.
- Copyright
- © 2025 The Author(s)
- Open Access
- Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
Cite this article
TY - CONF AU - R. Rajivkannan AU - V. Sharmila AU - M. Venkatesan AU - M. R. Avinesh AU - R. Avandivarma AU - K. Arunkumar PY - 2025 DA - 2025/05/23 TI - Deep Temporal Seizure Prediction A CNN-LSTM Hybrid for Enhanced EEG Signal Analysis BT - Proceedings of the International Conference on Sustainability Innovation in Computing and Engineering (ICSICE 2024) PB - Atlantis Press SP - 424 EP - 437 SN - 2352-538X UR - https://doi.org/10.2991/978-94-6463-718-2_37 DO - 10.2991/978-94-6463-718-2_37 ID - Rajivkannan2025 ER -