Ensemble Deep Learning Technique for Robust Seizure Detection ‎Using Integrating Convolutional and Recurrent Neural Networks ‎with Advanced Optimization Techniques

  • Authors

    • K. Dileep Kumar Ph.D. Scholar, Computer Science & Engineering, GIET University, Odisha
    • Dr. Sachikanta Dash Associate Professor, GIET University, Odisha
    • Dr. Rajendra Kumar Ganiya A. Department of Computer Science and Engineering,‎ Koneru Lakshmaiah Education Foundation, Vaddeswaram, Guntur
    https://doi.org/10.14419/04tp2d60

    Received date: July 13, 2025

    Accepted date: July 25, 2025

    Published date: November 1, 2025

  • Seizure Detection; EEG Signals; Deep Learning; CNN-LSTM Model; Adversarial Attacks; ‎Robustness; Hybrid Models; Temporal Features

  • Abstract

    Electroencephalogram (EEG) is a common problem where seizures need to be detected in the ‎healthcare industry because the quality of life of a patient can be affected, and it can even ‎cause serious health issues. In this paper, a hybrid deep learning model combining ‎Convolutional Neural Networks (CNN) to capture spatial features and Long Short-Term ‎Memory (LSTM) RNNs to detect temporal dynamics in EEG signals is presented. Our model ‎was trained and tested on the CHB-MIT Scalp EEG Database, which has been used in many ‎studies on seizure detection. The hybrid CNN-LSTM structure does better than the ‎conventional CNN or LSTM structure and its accuracy is 94.3 %, its precision score is 92.9 ‎and it has a high recall score of 99.2 with F1 score of around CNN-only model, however, ‎performed a little better gaining 89.5 percent of accuracy along with only approximately 86.6 ‎percent of increase in F1 score. The LSTM has also performed badly with a range of 84-86%. ‎Moreover, after conducting the study, the research also takes a step further in determining ‎how robust these models are against adversarial attack by attacking them with attacks using ‎Fast Gradient Sign Method (FGSM). At 0.05, the accuracy was reduced which is expected to ‎happen to all models but the hybrid model had a lesser impact hence returning with 78:5% ‎inaccuracy. The hybrid model (0.833) had better robustness score compared to the CNN-only ‎and the LSTM-only model as in the case of main model results with MaxFillup. Adversarial ‎training also plays a critical role in enhancing the underpinnings of this hybrid model such that ‎when it was tested on clean data, it was tested robustly against white-box attacks. The hybrid ‎model proposes a possible approach toward real-time seizure detection in clinics, which ‎besides being more accurate can be adversarailly robust as well. The contributes to the current ‎body of literature on seizure detection a fully integrated approach toward robust behavior in ‎practical medical environments‎.

  • References

    1. Ramanji, Rs, Archana Ca, Suresh Kp, and Naveesh Yb. "Revolutionizing Potato Crop Management: Deep Learning-Driven Potato Disease Detec-tion with Convolutional Neural Networks." International Journal of Advanced Biochemistry Research, p. 650, 2023.
    2. Ingolfsson, Thorir, et al. "EPIDENET: An Energy-Efficient Approach to Seizure Detection for Embedded Systems." IEEE, pp. 1-10, 2023. https://doi.org/10.1109/BioCAS58349.2023.10388554.
    3. Omari, Swaleh, et al. "Enhancing EEG Signals Classification Using LSTM-CNN Architecture." Jomo Kenyatta University of Agriculture and Technology, 21 June 2023. https://doi.org/10.22541/au.168737045.52367778/v1.
    4. Endah, Marselina, et al. "Journal of Soft Computing Exploration." Journal of Soft Computing Exploration, pp. 195-203, 2023. https://doi.org/10.52465/joscex.v4i4.226.
    5. Albattah, Albatul, and Murad Rassam. "Detection of Adversarial Attacks Against the Hybrid Convolutional Long Short-Term Memory Deep Learning Technique for Healthcare Monitoring Applications." Applied Sciences, vol. 13, 2023, p. 6807. https://doi.org/10.3390/app13116807.
    6. Debelo, Biniam, Bheema Lingaiah Thamineni, Hanumesh Dasari, and Ahmed Dawud. "Detection and Severity Identification of Neonatal Seizure Using Deep Convolutional Neural Networks from Multichannel EEG Signal." Pediatric Health, Medicine and Therapeutics, pp. 405-417, 2023.
    7. Mohammed, Widad, Mohammed Taha, and Haider Abduljabbar. "Deep Learning for Malaria Diagnosis: Leveraging Convolutional Neural Net-works for Accurate Parasite Detection." Journal of King Saud University-Computer and Information Sciences, pp. 1700-1705, 2022. https://doi.org/10.2147/PHMT.S427773.
    8. Franco, Nicola Rares, et al. "Deep Learning-Based Surrogate Models for Parametrized PDEs: Handling Geometric Variability Through Graph Neu-ral Networks." Journal/Conference Proceeding, pp. 1-3, Year 2022. https://doi.org/10.1063/5.0170101.
    9. Oie, Grant, et al. "Application of Long Short-Term Memory Deep Learning Networks on Very-High-Energy Gamma-Ray Classification with VER-ITAS." 38th International Cosmic Ray Conference (ICRC2023), 26 July - 3 August 2023, Nagoya, Japan. https://doi.org/10.22323/1.444.0692.
    10. Guo, Yinjing, et al. "An Epileptic Seizure Prediction Method Based on CBAM-3D CNN-LSTM Model." IEEE Journal of Translational Engineer-ing in Health and Medicine, 2023.
    11. Li, Y., et al. "Detection of Epileptic Seizures Using Hybrid Deep Learning Approaches." Journal/Conference Proceeding, Year 2022.
    12. Song, Z., et al. "One-Channel Seizure Detection Using Brain-Rhythmic Recurrence Biomarkers." Journal/Conference Proceeding, Year 2022. https://doi.org/10.1109/TNSRE.2022.3165060.
    13. Strohmeier, Christian Brodbeck, Lauri Parkkonen, and Matti S. Hamalainen. "MNE Software for Processing MEG and EEG Data." Neuroimage, vol. 86, 2013, pp. 446-460. https://doi.org/10.1016/j.neuroimage.2013.10.027.
    14. Strohmeier, Christian Brodbeck, Roman Goj, Mainak Jas, Teon Brooks, Lauri Parkkonen, and Matti S. Hamalainen. "MEG and EEG Data Analysis with MNE-Python." Frontiers in Neuroscience, vol. 7, no. 267, 2013, pp. 1-13. https://doi.org/10.3389/fnins.2013.00267.
    15. Sachikanta Dash, G Rajendra Kumar. "Detection of Epileptic Seizures Using Hybrid Deep Learning Approaches." International Journal of Intelli-gent Systems and Applications in Engineering, Year 2022.
    16. Kumar, K. Dileep, Sachikanta Dash, and Ganiya, Rajendra Kumar. "International Journal of Intelligent Systems and Applications in Engineering." Journal/Proceeding, 2023.

  • Downloads

  • How to Cite

    Kumar, K. D. ., Dash, D. S. ., & A., D. R. K. G. . (2025). Ensemble Deep Learning Technique for Robust Seizure Detection ‎Using Integrating Convolutional and Recurrent Neural Networks ‎with Advanced Optimization Techniques. International Journal of Basic and Applied Sciences, 14(SI-1), 510-518. https://doi.org/10.14419/04tp2d60