A generalizable epilepsy detection network based on dual-attention mechanism_Journal of Biomedical Engineering

Authors：

ZHANG Lei ¹ ,  ZHAI Fengwen ¹ , JIN Jing ¹ , JIANG Xiangde ² , CHANG Wenwen ¹

1. School of Electronic and Information Engineering, Lanzhou Jiaotong University, Lanzhou 730070, P. R. China;
2. Electronics Wanwei Information Technology Co. Ltd., Lanzhou 730030, P. R. China;

Corresponding?author：

ZHAI Fengwen, Email: zhaifw@mail.lzjtu.cn

Keywords：

Electroencephalogram; Epilepsy detection; Deep learning; Generalization; Feature extraction

DOI：

10.7507/1001-5515.202508053

Video：

Export PDF Favorites Scan Get Citation

Abstract Full text Figures/Tables Video References Cited by

Existing deep learning models for epileptic electroencephalogram (EEG) signal analysis frequently overlook intrinsic pathological characteristics during feature extraction and exhibit insufficient cross-dataset generalization. To address these limitations, this study proposes an innovative dual-attention epilepsy detection network (EDDANet). The model integrates a multi-band and multi-scale dual-attention module with a dynamic kernel sampling adaptive convolutional module to classify interictal and ictal EEG signals. Extensive experiments conducted on four heterogeneous public datasets demonstrate that EDDANet consistently outperforms state-of-the-art models across key evaluation metrics, including accuracy and recall. Notably, this work is the first to achieve robust generalization across varying lead configurations, sampling rates, and electrode layouts. In conclusion, this study provides a valuable methodological framework for the design and optimization of automated epilepsy detection systems in complex scenarios, providing reference for enhancing the generalizability and clinical utility of deep learning models in real-world environments.

Citation： ZHANG Lei, ZHAI Fengwen, JIN Jing, JIANG Xiangde, CHANG Wenwen. A generalizable epilepsy detection network based on dual-attention mechanism. Journal of Biomedical Engineering, 2026, 43(3): 530-538. doi: 10.7507/1001-5515.202508053 Copy

1.	Johnson E L. Seizures and epilepsy. Med Clin North Am, 2019, 103(2): 309-324.
2.	王興起, 李明愛. 應用基于注意力的多尺度殘差網絡實現癲癇腦電自動檢測. 生物醫學工程學雜志, 2024, 41(2): 253-261.
3.	Aboyeji S T, Adebiyi A A, Akinola S O, et al. A novel hybrid feature selection technique for epileptic seizure classification. IEEE Trans Hum Mach Syst, 2026, 56(2): 331-342.
4.	Zhang T, He L, Zhang D, et al. DAFF-SNN: dual attention-driven and feature fusion-based spiking neural network for epilepsy detection based on electroencephalogram. IEEE J Biomed Health Inform, 2026, 30(1): 315-327.
5.	Sun J, Cao Z, Zhao J. Multi-channel fusion deep wavelet spectrum network for epileptic signal classification. IEEE J Biomed Health Inform, 2026, 30(4): 3495-3506.
6.	Carzaniga F S, Hersche M, Schindler K A, et al. A composable channel-adaptive architecture for seizure classification. IEEE J Biomed Health Inform, 2026, 30(1): 279-291.
7.	Liao J, Liu Y, Wang H, et al. An EEG-based seizure prediction model encoding brain network temporal dynamics. IEEE J Biomed Health Inform, 2025, 29(11): 8059-8072.
8.	Dong Q, Zhang H, Xiao J, et al. Multi-scale spatio-temporal attention network for epileptic seizure prediction. IEEE J Biomed Health Inform, 2025, 29(7): 4784-4795.
9.	Lian Q, Wang Y, Qi Y. Dynamic instance-level graph learning network of intracranial electroencephalography signals for epileptic seizure prediction. IEEE J Biomed Health Inform, 2025, 29(11): 8348-8360.
10.	周嵩, 高天寒. 基于注意力機制RNN模型的癲癇患者腦電信號識別方法. 東北大學學報(自然科學版), 2023, 44(8): 1098-1103.
11.	袁月婷. 基于腦電信號多域特征的駕駛行為識別研究. 蘭州: 蘭州交通大學, 2024.
12.	Niestroj L M, Perez-Palma E, Howrigan D P, et al. Epilepsy subtype-specific copy number burden observed in a genome-wide study of 17 458 subjects. Brain, 2020, 143(7): 2106-2118.
13.	Thijs R D, Surges R, O'Brien T J, et al. Epilepsy in adults. Lancet, 2019, 393(10172): 689-701.
14.	何曉明. 基于CNN-BiGRU的癲癇腦電信號識別研究. 科技創新與生產力, 2025, 46(10): 124-127.
15.	Bai X, Wang R, Yang J, et al. Exploratory study on the dynamic characteristics of brain electrical impedance during epileptic seizures. IEEE Sens J, 2025, 25(12): 22841-22849.
16.	Vaddi V N, Kodali P. Enhanced epilepsy sensitivity and detection rate with improved specificity by integration of modified LSTM networks. IEEE Sens J, 2025, 25(7): 11963-11970.
17.	韓長明, 彭福來, 陳財, 等. 基于腦電信號的癲癇發作預測研究進展. 生物醫學工程學雜志, 2021, 38(6): 1193-1202.
18.	張漢明, 馬金剛, 張寧寧, 等. 深度學習在癲癇檢測中的應用進展. 計算機工程與應用, 2023, 59(10): 35-47.
19.	徐晴, 葛成, 蔡標, 等. 基于深度學習的癲癇腦電信號分類. 數據采集與處理, 2022, 37(4): 787-797.
20.	Lawhern V J, Solon A J, Waytowich N R, et al. EEGNet: a compact convolutional neural network for EEG-based brain-computer interfaces. J Neural Eng, 2018, 15(5): 056013.
21.	Thuwajit P, Rangpong P, Sawangjai P, et al. EEGWaveNet: multiscale CNN-based spatiotemporal feature extraction for EEG seizure detection. IEEE Trans Ind Inform, 2022, 18(8): 5547-5557.
22.	Lim Z Y, Pang Y H, Ooi S Y, et al. MLTCN-EEG: metric learning-based temporal convolutional network for seizure EEG classification. Neural Comput Appl, 2025, 37: 2849-2875.
23.	Wang X, Wang G, Wu T, et al. EEG-based seizure onset detection of frontal and temporal lobe epilepsies using 1DCNN. IEEE Trans Neural Syst Rehabil Eng, 2025, 33: 2263-2272.
24.	梁國華, 額吉納, 李涵懿, 等. 基于圖注意力網絡的癲癇發作異常檢測模型. 生物醫學工程學雜志, 2025, 42(4): 693-700.
25.	Detti P, Vatti G, de Lara G Z M, et al. EEG synchronization analysis for seizure prediction: a study on data of noninvasive recordings. Processes, 2020, 8(7): 846.
26.	Andrzejak R G, Lehnertz K, Rieke C, et al. Indications of nonlinear deterministic and finite dimensional structures in time series of brain electrical activity: dependence on recording region and brain state. Phys Rev E, 2001, 64: 061907.
27.	Burrello A, Schindler K, Benini L, et al. Hyperdimensional computing with local binary patterns: one-shot learning of seizure onset and identification of ictogenic brain regions using short-time iEEG recordings. IEEE Trans Biomed Eng, 2020, 67(2): 601-613.
28.	Sun S, Li J, Chen H, et al. A study of resting-state EEG biomarkers for depression recognition. arXiv, 2022: 2002.11039.
29.	張培玲, 候康. 基于多尺度自適應殘差網絡的癲癇檢測方法. 中國醫學物理學雜志, 2025, 42(3): 381-387.
30.	Inouye T, Toi S, Matsumoto Y. A new segmentation method of electroencephalograms by use of Akaike's information criterion. Cogn Brain Res, 1995, 3(1): 33-40.
31.	Gajic D, Djurovic Z, Gligorijevic J, et al. Detection of epileptiform activity in EEG signals based on timefrequency and non-linear analysis. Front Comput Neurosci, 2015, 9: 38.
32.	常禧龍, 梁琨, 李文濤. 深度學習優化器進展綜述. 計算機工程與應用, 2024, 60(7): 1-12.
33.	Sun F, Zhai F, Jin J. Depression EEG signal classification combining temporal convolution and feature calibration. J Meas Sci Instrum, 2025, 16(4): 547-557.
34.	Alotaibi Y, Al-Luhaybi M, Alharbi J, et al. Hybrid CNN-TCN modeling of EEG temporal signals for epileptic seizure prediction//2025 International Conference on Advanced Machine Learning and Data Science (AMLDS), Tokyo: IEEE, 2025: 371-377.
35.	Maaten L, Hinton G. Visualizing data using t-SNE. J Mach Learn Res, 2008, 9: 2579-2605.
36.	Song Y, Zheng Q, Liu B, et al. EEG conformer: convolutional transformer for EEG decoding and visualization. IEEE Trans Neural Syst Rehabil Eng, 2023, 31: 710-719.
37.	Ma Y, Huang Z, Su J, et al. A multi-channel feature fusion CNN-Bi-LSTM epilepsy EEG classification and prediction model based on attention mechanism. IEEE Access, 2023, 11: 62855-62864.
38.	Zhao W, Wang W, Patnaik L, et al. Residual and bidirectional LSTM for epileptic seizure detection. Front Comput Neurosci, 2024, 18: 1415967.
39.	Khalfallah S, Bouallegue K. Novel enhanced ChronoNet model for epileptic seizure detection: a comprehensive methodological study//2025 IEEE 22nd International Multi-Conference on Systems, Signals & Devices (SSD), Monastir: IEEE, 2025: 1161-1168.
40.	王帥. 基于深度可遷移性特征學習的跨病人癲癇發作檢測研究. 濟南: 山東師范大學, 2025.

1. Johnson E L. Seizures and epilepsy. Med Clin North Am, 2019, 103(2): 309-324.
2. 王興起, 李明愛. 應用基于注意力的多尺度殘差網絡實現癲癇腦電自動檢測. 生物醫學工程學雜志, 2024, 41(2): 253-261.
3. Aboyeji S T, Adebiyi A A, Akinola S O, et al. A novel hybrid feature selection technique for epileptic seizure classification. IEEE Trans Hum Mach Syst, 2026, 56(2): 331-342.
4. Zhang T, He L, Zhang D, et al. DAFF-SNN: dual attention-driven and feature fusion-based spiking neural network for epilepsy detection based on electroencephalogram. IEEE J Biomed Health Inform, 2026, 30(1): 315-327.
5. Sun J, Cao Z, Zhao J. Multi-channel fusion deep wavelet spectrum network for epileptic signal classification. IEEE J Biomed Health Inform, 2026, 30(4): 3495-3506.
6. Carzaniga F S, Hersche M, Schindler K A, et al. A composable channel-adaptive architecture for seizure classification. IEEE J Biomed Health Inform, 2026, 30(1): 279-291.
7. Liao J, Liu Y, Wang H, et al. An EEG-based seizure prediction model encoding brain network temporal dynamics. IEEE J Biomed Health Inform, 2025, 29(11): 8059-8072.
8. Dong Q, Zhang H, Xiao J, et al. Multi-scale spatio-temporal attention network for epileptic seizure prediction. IEEE J Biomed Health Inform, 2025, 29(7): 4784-4795.
9. Lian Q, Wang Y, Qi Y. Dynamic instance-level graph learning network of intracranial electroencephalography signals for epileptic seizure prediction. IEEE J Biomed Health Inform, 2025, 29(11): 8348-8360.
10. 周嵩, 高天寒. 基于注意力機制RNN模型的癲癇患者腦電信號識別方法. 東北大學學報(自然科學版), 2023, 44(8): 1098-1103.
11. 袁月婷. 基于腦電信號多域特征的駕駛行為識別研究. 蘭州: 蘭州交通大學, 2024.
12. Niestroj L M, Perez-Palma E, Howrigan D P, et al. Epilepsy subtype-specific copy number burden observed in a genome-wide study of 17 458 subjects. Brain, 2020, 143(7): 2106-2118.
13. Thijs R D, Surges R, O'Brien T J, et al. Epilepsy in adults. Lancet, 2019, 393(10172): 689-701.
14. 何曉明. 基于CNN-BiGRU的癲癇腦電信號識別研究. 科技創新與生產力, 2025, 46(10): 124-127.
15. Bai X, Wang R, Yang J, et al. Exploratory study on the dynamic characteristics of brain electrical impedance during epileptic seizures. IEEE Sens J, 2025, 25(12): 22841-22849.
16. Vaddi V N, Kodali P. Enhanced epilepsy sensitivity and detection rate with improved specificity by integration of modified LSTM networks. IEEE Sens J, 2025, 25(7): 11963-11970.
17. 韓長明, 彭福來, 陳財, 等. 基于腦電信號的癲癇發作預測研究進展. 生物醫學工程學雜志, 2021, 38(6): 1193-1202.
18. 張漢明, 馬金剛, 張寧寧, 等. 深度學習在癲癇檢測中的應用進展. 計算機工程與應用, 2023, 59(10): 35-47.
19. 徐晴, 葛成, 蔡標, 等. 基于深度學習的癲癇腦電信號分類. 數據采集與處理, 2022, 37(4): 787-797.
20. Lawhern V J, Solon A J, Waytowich N R, et al. EEGNet: a compact convolutional neural network for EEG-based brain-computer interfaces. J Neural Eng, 2018, 15(5): 056013.
21. Thuwajit P, Rangpong P, Sawangjai P, et al. EEGWaveNet: multiscale CNN-based spatiotemporal feature extraction for EEG seizure detection. IEEE Trans Ind Inform, 2022, 18(8): 5547-5557.
22. Lim Z Y, Pang Y H, Ooi S Y, et al. MLTCN-EEG: metric learning-based temporal convolutional network for seizure EEG classification. Neural Comput Appl, 2025, 37: 2849-2875.
23. Wang X, Wang G, Wu T, et al. EEG-based seizure onset detection of frontal and temporal lobe epilepsies using 1DCNN. IEEE Trans Neural Syst Rehabil Eng, 2025, 33: 2263-2272.
24. 梁國華, 額吉納, 李涵懿, 等. 基于圖注意力網絡的癲癇發作異常檢測模型. 生物醫學工程學雜志, 2025, 42(4): 693-700.
25. Detti P, Vatti G, de Lara G Z M, et al. EEG synchronization analysis for seizure prediction: a study on data of noninvasive recordings. Processes, 2020, 8(7): 846.
26. Andrzejak R G, Lehnertz K, Rieke C, et al. Indications of nonlinear deterministic and finite dimensional structures in time series of brain electrical activity: dependence on recording region and brain state. Phys Rev E, 2001, 64: 061907.
27. Burrello A, Schindler K, Benini L, et al. Hyperdimensional computing with local binary patterns: one-shot learning of seizure onset and identification of ictogenic brain regions using short-time iEEG recordings. IEEE Trans Biomed Eng, 2020, 67(2): 601-613.
28. Sun S, Li J, Chen H, et al. A study of resting-state EEG biomarkers for depression recognition. arXiv, 2022: 2002.11039.
29. 張培玲, 候康. 基于多尺度自適應殘差網絡的癲癇檢測方法. 中國醫學物理學雜志, 2025, 42(3): 381-387.
30. Inouye T, Toi S, Matsumoto Y. A new segmentation method of electroencephalograms by use of Akaike's information criterion. Cogn Brain Res, 1995, 3(1): 33-40.
31. Gajic D, Djurovic Z, Gligorijevic J, et al. Detection of epileptiform activity in EEG signals based on timefrequency and non-linear analysis. Front Comput Neurosci, 2015, 9: 38.
32. 常禧龍, 梁琨, 李文濤. 深度學習優化器進展綜述. 計算機工程與應用, 2024, 60(7): 1-12.
33. Sun F, Zhai F, Jin J. Depression EEG signal classification combining temporal convolution and feature calibration. J Meas Sci Instrum, 2025, 16(4): 547-557.
34. Alotaibi Y, Al-Luhaybi M, Alharbi J, et al. Hybrid CNN-TCN modeling of EEG temporal signals for epileptic seizure prediction//2025 International Conference on Advanced Machine Learning and Data Science (AMLDS), Tokyo: IEEE, 2025: 371-377.
35. Maaten L, Hinton G. Visualizing data using t-SNE. J Mach Learn Res, 2008, 9: 2579-2605.
36. Song Y, Zheng Q, Liu B, et al. EEG conformer: convolutional transformer for EEG decoding and visualization. IEEE Trans Neural Syst Rehabil Eng, 2023, 31: 710-719.
37. Ma Y, Huang Z, Su J, et al. A multi-channel feature fusion CNN-Bi-LSTM epilepsy EEG classification and prediction model based on attention mechanism. IEEE Access, 2023, 11: 62855-62864.
38. Zhao W, Wang W, Patnaik L, et al. Residual and bidirectional LSTM for epileptic seizure detection. Front Comput Neurosci, 2024, 18: 1415967.
39. Khalfallah S, Bouallegue K. Novel enhanced ChronoNet model for epileptic seizure detection: a comprehensive methodological study//2025 IEEE 22nd International Multi-Conference on Systems, Signals & Devices (SSD), Monastir: IEEE, 2025: 1161-1168.
40. 王帥. 基于深度可遷移性特征學習的跨病人癲癇發作檢測研究. 濟南: 山東師范大學, 2025.

Journal of Biomedical Engineering

A generalizable epilepsy detection network based on dual-attention mechanism

Abstract Full text Figures/Tables Video References Cited by

Previous Article

Next Article

Format

Content