The Journal of the Korea institute of electronic communication sciences (한국전자통신학회논문지)

Korea Institute of Electronic Communication Science (한국전자통신학회)
권호 표지
DOI QR코드

DOI QR Code

Analysis and Study for Appropriate Deep Neural Network Structures and Self-Supervised Learning-based Brain Signal Data Representation Methods

딥 뉴럴 네트워크의 적절한 구조 및 자가-지도 학습 방법에 따른 뇌신호 데이터 표현 기술 분석 및 고찰

  • Won-Jun Ko (School of AI Convergence, Sungshin Women's University)
  • 고원준 (성신여자대학교 AI융합학부)
  • Received : 2023.12.28
  • Accepted : 2024.02.17
  • Published : 2024.02.29
Download PDF
⟨ Previous Next ⟩

Abstract

Recently, deep learning technology has become those methods as de facto standards in the area of medical data representation. But, deep learning inherently requires a large amount of training data, which poses a challenge for its direct application in the medical field where acquiring large-scale data is not straightforward. Additionally, brain signal modalities also suffer from these problems owing to the high variability. Research has focused on designing deep neural network structures capable of effectively extracting spectro-spatio-temporal characteristics of brain signals, or employing self-supervised learning methods to pre-learn the neurophysiological features of brain signals. This paper analyzes methodologies used to handle small-scale data in emerging fields such as brain-computer interfaces and brain signal-based state prediction, presenting future directions for these technologies. At first, this paper examines deep neural network structures for representing brain signals, then analyzes self-supervised learning methodologies aimed at efficiently learning the characteristics of brain signals. Finally, the paper discusses key insights and future directions for deep learning-based brain signal analysis.

최근, 의료 데이터 표현 분야에서 딥러닝 방법들이 사실상의 표준으로 자리잡고 있다. 하지만, 딥러닝 기술은 내재적으로 많은 양의 학습 데이터를 필요로 하므로 대규모의 데이터를 확보하기 쉽지 않은 의료 분야에서는 직접적인 적용이 어려운 실정이다. 특히 뇌신호 모달리티의 경우, 변동성이 크기 때문에 여전히 데이터 부족 문제를 가진다. 이에, 최근 연구에서는 뇌신호의 시간-공간-주파수 특징을 적절하게 추출할 수 있는 딥 뉴럴 네트워크 구조를 설계하거나, 혹은 자가-지도 학습 방법을 도입하여 뇌신호의 신경생리학적 특징을 미리 학습하도록 한다. 본 논문에서는, 최근 각광받는 기술인 뇌-컴퓨터 인터페이스 및 피험자 상태 예측 등의 관점에서 소규모데이터를 다루기 위해 적용되는 방법론에 대한 분석 및 향후 기술 방향성을 제시한다. 먼저 현재 제안되고 있는 뇌신호 표현을 위한 딥 뉴럴 네트워크 구조에 대해 분석한다. 또한 뇌신호의 특성을 잘 학습하기 위한 자가-지도 학습 방법론을 분석한다. 끝으로, 딥러닝 기반 뇌신호 분석을 위한 중요 시사점 및 방향성에 관하여 논한다.

Keywords

References

  1. W. Ko, E. Jeon, S. Jeong, and H. Suk, "Multi-scale neural network for EEG representation learning in BCI," IEEE Comput. Intell. Mag., vol. 16, no. 2, 2021, pp. 31-45. https://doi.org/10.1109/MCI.2021.3061875
  2. C. Yang, D. Xiao, M. B. Westover, and J. Sun, "Self-supervised EEG representation learning for automatic sleep staging," arXiv, vol. 1, 2021.
  3. M. N. Mohsenvand, M. R. Izadi, and P. Maes, "Contrastive representation learning for electroencephalogram classification," in Proc. Mach. Learn. Health. Virtual Meeting, 2020.
  4. R. T. Schirrmeister, J. T. Springenberg, L. D. J. Fiederer, M. Glasstetter, K. Eggensperger, M. Tangermann, F. Hutter, W. Burgard, and T. Ball, "Deep learning with convolutional neural networks for EEG decoding and visualization," Hum. Brain Mapp., vol. 38, no. 11, 2017, pp. 5391-5420. https://doi.org/10.1002/hbm.23730
  5. D. Kim, S. Park, and D. Kim, "The classification scheme of ADHD for children based on the CNN model," J. of the Korea Institute of Electronic Communication Sciences, vol. 17, no. 5, 2022, pp. 809-814.
  6. V. J. Lawhern, A. J. Solon, N. R. Waytowich, S. M. Gordon, C. P. Hung, and B. J. Lance, "EEGNet: a compact convolutional neural network for EEG-based brain-computer interface," J. Neural Eng., vol. 15, no. 5, 2018, pp. 056013.
  7. E. Santamaria-Vazquez, V. Martinez-Cagigal, F. Vaquerizo-Villar, and R. Hornero, "EEG-inception: a novel deep convolutional nerual network for assistive ERP-based brain-computer interfaces," IEEE Trans. Neural Syst. Rehabil. Eng., vol. 28, no. 12, 2020, pp. 2773-2782. https://doi.org/10.1109/TNSRE.2020.3048106
  8. E. Jeon, W. Ko, J. Yoon, and H. Suk, "Mutual information-driven subject-invariant and class-relevant deep representation learning in BCI," IEEE Trans. Neural Net. Learn. Syst., vol. 34, no. 2, 2021, pp. 739-749. https://doi.org/10.1109/TNNLS.2021.3100583
  9. O. Kwon, M. Lee, C. Guan, and S. Lee, "Subject-independent brain-computer interfaces based on deep convolutional neural networks," IEEE Trans. Nerual Netw. Learn. Syst., vol. 31, no. 10, 2019, pp. 3839-3852. https://doi.org/10.1109/TNNLS.2019.2946869
  10. H. Banville, O. Chehab, A. Hyvarinen, D.-A. Engemann, and A. Gramfort, "Uncovering the structure of clinical EEG signals with self-supervised learning," J. Neural Eng., vol. 18, no. 4, 2021, p. 046020.
  11. X. Jaing, J. Zhao, B. Du, and Z. Yuan, "Self-supervised contrastive learning for EEG-based sleep staging," in Proc. Int. Joint Conf. Neural Netw., Virtual Meeting, 2021.
  12. Q. Xiao, J. Wang, J. Ye, H. Zhang, Y. Bu, Y. Zhang, and H. Wu, "Self-supervised learning for sleep stage classification with predictive and discriminative contrastive coding," in Proc. IEEE Int. Conf. Acoust., Speech, Signal Process, Toronto, Canada, 2021.
  13. S. Perez-Velasco, E. Santamaria-Vazquez, V. Martinez-Cagigal, D. Marcos-Martinez, and R. Hornero, "EEGSym: overcoming inter-subject variability in motor imagery based BCIs with deep learning," IEEE Trans. Neural Syst. Rehabil. Eng., vol. 30, 2022, pp. 1766-1775. https://doi.org/10.1109/TNSRE.2022.3186442
  14. W. Ko and H. Suk, "EEG-oriented self-supervised learning and cluster-aware adaptation," in Proc. Int. ACM Conf. Inf. Knowl. Manag., Atlanta, USA, 2022.
  15. Y. ou, S. Sun, H. Gan, R. Zhou, and Z. Yang, "An improved self-supervised learning for EEG classification," Math. Biosci. Eng., vol. 19, no. 7, 2022, pp. 6907-6922. https://doi.org/10.3934/mbe.2022325
  16. S. Tang, J. A. Dunnmon, K. Saab, X. Zhang, Q. Huang, E. Dubost, D. L. Rubin, and C. Lee-Messer, "Self-supervised graph neural networks for improved electroencephalographic seizure analysis," arXiv, vol. 1, 2021.
  17. J. Moon and Y. Lee, "Artificial Intelligence Computing Platform Design for Underwater Localization," J. of the Korea Institute of Electronic Communication Sciences, vol. 17, no. 1, 2022, pp. 119-124.