Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Aug 21;11(1):21.
doi: 10.1186/s40708-024-00234-x.

HyEpiSeiD: a hybrid convolutional neural network and gated recurrent unit model for epileptic seizure detection from electroencephalogram signals

Rajdeep Bhadra  1 , Pawan Kumar Singh  1   2 , Mufti Mahmud  3   4   5
Affiliations

HyEpiSeiD: a hybrid convolutional neural network and gated recurrent unit model for epileptic seizure detection from electroencephalogram signals

Rajdeep Bhadra et al. Brain Inform. .

Abstract

Epileptic seizure (ES) detection is an active research area, that aims at patient-specific ES detection with high accuracy from electroencephalogram (EEG) signals. The early detection of seizure is crucial for timely medical intervention and prevention of further injuries of the patients. This work proposes a robust deep learning framework called HyEpiSeiD that extracts self-trained features from the pre-processed EEG signals using a hybrid combination of convolutional neural network followed by two gated recurrent unit layers and performs prediction based on those extracted features. The proposed HyEpiSeiD framework is evaluated on two public datasets, the UCI Epilepsy and Mendeley datasets. The proposed HyEpiSeiD model achieved 99.01% and 97.50% classification accuracy, respectively, outperforming most of the state-of-the-art methods in epilepsy detection domain.

Keywords: Convolutional neural network; Electroencephalogram signals; Epilepsy UCI dataset; Epileptic seizure detection; Gated recurrent unit; HyEpiSeiD; Mendeley dataset.

PubMed Disclaimer

Conflict of interest statement

Mufti Mahmud is an editorial board member of the Brain Informatics journal and was not involved in the editorial review or the decision to publish this article. The authors declare no Competing interests.

Figures

Fig. 1
Fig. 1
Pictorial Representation of different phases of our model pipeline for epilepsy seizure detection from raw EEG data
Fig. 2
Fig. 2
Detailed Architecture of our proposed HyEpiSeiD framework for Epilepsy seizure detection from EEG signals
Fig. 3
Fig. 3
General architecture of an GRU recurrent neural network.The boxes denote nodes, the upward arrow indicate the previous input and the forward arrow indicates passing the output to the next node
Fig. 4
Fig. 4
Confusion matrix of our proposed HyEpiSeiD model on UCI Epilepsy dataset for 2-class classification problem
Fig. 5
Fig. 5
Confusion matrix of our proposed HyEpiSeiD model on UCI Epilepsy dataset for 5-class classification problem
Fig. 6
Fig. 6
loss vs epoch and accuracy vs epoch curve of our proposed HyEpiSeiD model for UCI Epilepsy dataset in 2-class classification problem
Fig. 7
Fig. 7
Loss vs epoch and accuracy vs epoch curve of our proposed HyEpiSeiD model on UCI Epilepsy dataset for 5-class classification problem
Fig. 8
Fig. 8
Bar chart showing the class-wise metric scores of our proposed HyEpiSeiD model for 2-class classification on UCI Epilepsy dataset
Fig. 9
Fig. 9
Bar chart showing the class-wise metric scores of our proposed HyEpiSeiD model for 5-class classification on UCI Epilepsy dataset
Fig. 10
Fig. 10
Confusion matrix of our proposed HyEpiSeiD model on Mendeley dataset for 2-class classification
Fig. 11
Fig. 11
Loss vs epoch and accuracy vs epoch curve of our proposed HyEpiSeiD model on Mendeley dataset for 2-class classification
Fig. 12
Fig. 12
Bar chart of class-wise metric scores of our proposed HyEpiSeiD model for 2-class classification on Mendeley dataset
See this image and copyright information in PMC

References

    1. Abraira L, Gramegna LL, Quintana M, Santamarina E, Salas-Puig J, Sarria S, Rovira A, Toledo M (2019) Cerebrovascular disease burden in late-onset non-lesional focal epilepsy. Seizure 66:31–35 10.1016/j.seizure.2019.02.004 - DOI - PubMed
    1. Galanopoulou AS, Buckmaster PS, Staley KJ, Moshé SL, Perucca E, Engel J Jr, Löscher W, Noebels JL, Pitkänen A, Stables J et al (2012) Identification of new epilepsy treatments: issues in preclinical methodology. Epilepsia 53(3):571–582 10.1111/j.1528-1167.2011.03391.x - DOI - PMC - PubMed
    1. San-Segundo R, Gil-Martín M, D’Haro-Enríquez LF, Pardo JM (2019) Classification of epileptic EEG recordings using signal transforms and convolutional neural networks. Comput Biol Med 109:148–158 10.1016/j.compbiomed.2019.04.031 - DOI - PubMed
    1. Banerjee A, Sarkar A, Roy S, Singh PK, Sarkar R (2022) Covid-19 chest x-ray detection through blending ensemble of CNN snapshots. Biomed Signal Process Control 78:104000 10.1016/j.bspc.2022.104000 - DOI - PMC - PubMed
    1. Nashef L, Fish D, Garner S, Sander J, Shorvon S (1995) Sudden death in epilepsy: a study of incidence in a young cohort with epilepsy and learning difficulty. Epilepsia 36(12):1187–1194 10.1111/j.1528-1157.1995.tb01061.x - DOI - PubMed

LinkOut - more resources