Online Adaptive CNN: A Session-to-session Transfer Learning Approach for Non-stationary EEG

Shuailei Zhang; Dezhi Zheng; Ning Tang; Effie Chew; Rosary Yuting Lim; Kai Keng Ang; Cuntai Guan

doi:10.1109/SSCI51031.2022.10022274

Online Adaptive CNN: A Session-to-session Transfer Learning Approach for Non-stationary EEG

Shuailei Zhang, Dezhi Zheng, Ning Tang, Effie Chew, Rosary Yuting Lim, Kai Keng Ang, Cuntai Guan

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

7 Citations (Scopus)

Abstract

The convolutional neural network (CNN) automatically learns EEG representations in higher and nonlinear space via backpropagation and outputs the predictions in an end-to-end manner. Owing to these advantages, CNN has been used to decode electroencephalogram (EEG) and drive brain computer interface (BCI). However, its applications in BCI-assisted post-stroke neurorehabilitation remain limited for it is unable to address the inherent session-to-session non-stationarity in the EEG between the initial calibration session and subsequent online sessions. In this paper, we present a simple but effective online adaptive CNN (aCNN) to address the non-stationarity in multi-session EEG by progressively updating the subject-specific model. The performance of the proposed aCNN is evaluated on two neurorehabilitation datasets with a large population of post-stroke patients (33 patients with a total of 358 EEG sessions). Results indicate that, our proposed aCNN reaches at least as good a performance as the widely used online adaptive Filter Bank Common Spatial Patterns (aFBCSP) and with significantly higher accuracies than that for DeepConv and offline FBCSP algorithms. Our results support, for the first time, the use of a CNN-based adaptive learning method to decode non-stationary EEG signals for BCI-assisted post-stroke rehabilitation.

Original language	English
Title of host publication	Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022
Editors	Hisao Ishibuchi, Chee-Keong Kwoh, Ah-Hwee Tan, Dipti Srinivasan, Chunyan Miao, Anupam Trivedi, Keeley Crockett
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	164-170
Number of pages	7
ISBN (Electronic)	9781665487689
DOIs	https://doi.org/10.1109/SSCI51031.2022.10022274
Publication status	Published - 2022
Externally published	Yes
Event	2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022 - Singapore, Singapore Duration: 4 Dec 2022 → 7 Dec 2022

Publication series

Name	Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022

Conference

Conference	2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022
Country/Territory	Singapore
City	Singapore
Period	4/12/22 → 7/12/22

Keywords

brain computer interface
neurorehabilitation
non-stationarity
online adaptive CNN

Access to Document

10.1109/SSCI51031.2022.10022274

Cite this

Zhang, S., Zheng, D., Tang, N., Chew, E., Lim, R. Y., Ang, K. K., & Guan, C. (2022). Online Adaptive CNN: A Session-to-session Transfer Learning Approach for Non-stationary EEG. In H. Ishibuchi, C.-K. Kwoh, A.-H. Tan, D. Srinivasan, C. Miao, A. Trivedi, & K. Crockett (Eds.), Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022 (pp. 164-170). (Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SSCI51031.2022.10022274

Zhang, Shuailei ; Zheng, Dezhi ; Tang, Ning et al. / Online Adaptive CNN : A Session-to-session Transfer Learning Approach for Non-stationary EEG. Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022. editor / Hisao Ishibuchi ; Chee-Keong Kwoh ; Ah-Hwee Tan ; Dipti Srinivasan ; Chunyan Miao ; Anupam Trivedi ; Keeley Crockett. Institute of Electrical and Electronics Engineers Inc., 2022. pp. 164-170 (Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022).

@inproceedings{b30a3f2aac8b477ba171babc0054f4b0,

title = "Online Adaptive CNN: A Session-to-session Transfer Learning Approach for Non-stationary EEG",

abstract = "The convolutional neural network (CNN) automatically learns EEG representations in higher and nonlinear space via backpropagation and outputs the predictions in an end-to-end manner. Owing to these advantages, CNN has been used to decode electroencephalogram (EEG) and drive brain computer interface (BCI). However, its applications in BCI-assisted post-stroke neurorehabilitation remain limited for it is unable to address the inherent session-to-session non-stationarity in the EEG between the initial calibration session and subsequent online sessions. In this paper, we present a simple but effective online adaptive CNN (aCNN) to address the non-stationarity in multi-session EEG by progressively updating the subject-specific model. The performance of the proposed aCNN is evaluated on two neurorehabilitation datasets with a large population of post-stroke patients (33 patients with a total of 358 EEG sessions). Results indicate that, our proposed aCNN reaches at least as good a performance as the widely used online adaptive Filter Bank Common Spatial Patterns (aFBCSP) and with significantly higher accuracies than that for DeepConv and offline FBCSP algorithms. Our results support, for the first time, the use of a CNN-based adaptive learning method to decode non-stationary EEG signals for BCI-assisted post-stroke rehabilitation.",

keywords = "brain computer interface, neurorehabilitation, non-stationarity, online adaptive CNN",

author = "Shuailei Zhang and Dezhi Zheng and Ning Tang and Effie Chew and Lim, {Rosary Yuting} and Ang, {Kai Keng} and Cuntai Guan",

note = "Publisher Copyright: {\textcopyright} 2022 IEEE.; 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022 ; Conference date: 04-12-2022 Through 07-12-2022",

year = "2022",

doi = "10.1109/SSCI51031.2022.10022274",

language = "English",

series = "Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "164--170",

editor = "Hisao Ishibuchi and Chee-Keong Kwoh and Ah-Hwee Tan and Dipti Srinivasan and Chunyan Miao and Anupam Trivedi and Keeley Crockett",

booktitle = "Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022",

address = "United States",

}

Zhang, S, Zheng, D, Tang, N, Chew, E, Lim, RY, Ang, KK & Guan, C 2022, Online Adaptive CNN: A Session-to-session Transfer Learning Approach for Non-stationary EEG. in H Ishibuchi, C-K Kwoh, A-H Tan, D Srinivasan, C Miao, A Trivedi & K Crockett (eds), Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022. Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022, Institute of Electrical and Electronics Engineers Inc., pp. 164-170, 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022, Singapore, Singapore, 4/12/22. https://doi.org/10.1109/SSCI51031.2022.10022274

Online Adaptive CNN: A Session-to-session Transfer Learning Approach for Non-stationary EEG. / Zhang, Shuailei; Zheng, Dezhi; Tang, Ning et al.
Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022. ed. / Hisao Ishibuchi; Chee-Keong Kwoh; Ah-Hwee Tan; Dipti Srinivasan; Chunyan Miao; Anupam Trivedi; Keeley Crockett. Institute of Electrical and Electronics Engineers Inc., 2022. p. 164-170 (Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Online Adaptive CNN

T2 - 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022

AU - Zhang, Shuailei

AU - Zheng, Dezhi

AU - Tang, Ning

AU - Chew, Effie

AU - Lim, Rosary Yuting

AU - Ang, Kai Keng

AU - Guan, Cuntai

PY - 2022

Y1 - 2022

N2 - The convolutional neural network (CNN) automatically learns EEG representations in higher and nonlinear space via backpropagation and outputs the predictions in an end-to-end manner. Owing to these advantages, CNN has been used to decode electroencephalogram (EEG) and drive brain computer interface (BCI). However, its applications in BCI-assisted post-stroke neurorehabilitation remain limited for it is unable to address the inherent session-to-session non-stationarity in the EEG between the initial calibration session and subsequent online sessions. In this paper, we present a simple but effective online adaptive CNN (aCNN) to address the non-stationarity in multi-session EEG by progressively updating the subject-specific model. The performance of the proposed aCNN is evaluated on two neurorehabilitation datasets with a large population of post-stroke patients (33 patients with a total of 358 EEG sessions). Results indicate that, our proposed aCNN reaches at least as good a performance as the widely used online adaptive Filter Bank Common Spatial Patterns (aFBCSP) and with significantly higher accuracies than that for DeepConv and offline FBCSP algorithms. Our results support, for the first time, the use of a CNN-based adaptive learning method to decode non-stationary EEG signals for BCI-assisted post-stroke rehabilitation.

AB - The convolutional neural network (CNN) automatically learns EEG representations in higher and nonlinear space via backpropagation and outputs the predictions in an end-to-end manner. Owing to these advantages, CNN has been used to decode electroencephalogram (EEG) and drive brain computer interface (BCI). However, its applications in BCI-assisted post-stroke neurorehabilitation remain limited for it is unable to address the inherent session-to-session non-stationarity in the EEG between the initial calibration session and subsequent online sessions. In this paper, we present a simple but effective online adaptive CNN (aCNN) to address the non-stationarity in multi-session EEG by progressively updating the subject-specific model. The performance of the proposed aCNN is evaluated on two neurorehabilitation datasets with a large population of post-stroke patients (33 patients with a total of 358 EEG sessions). Results indicate that, our proposed aCNN reaches at least as good a performance as the widely used online adaptive Filter Bank Common Spatial Patterns (aFBCSP) and with significantly higher accuracies than that for DeepConv and offline FBCSP algorithms. Our results support, for the first time, the use of a CNN-based adaptive learning method to decode non-stationary EEG signals for BCI-assisted post-stroke rehabilitation.

KW - brain computer interface

KW - neurorehabilitation

KW - non-stationarity

KW - online adaptive CNN

UR - http://www.scopus.com/inward/record.url?scp=85147796088&partnerID=8YFLogxK

U2 - 10.1109/SSCI51031.2022.10022274

DO - 10.1109/SSCI51031.2022.10022274

M3 - Conference contribution

AN - SCOPUS:85147796088

T3 - Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022

SP - 164

EP - 170

BT - Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022

A2 - Ishibuchi, Hisao

A2 - Kwoh, Chee-Keong

A2 - Tan, Ah-Hwee

A2 - Srinivasan, Dipti

A2 - Miao, Chunyan

A2 - Trivedi, Anupam

A2 - Crockett, Keeley

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 4 December 2022 through 7 December 2022

ER -

Zhang S, Zheng D, Tang N, Chew E, Lim RY, Ang KK et al. Online Adaptive CNN: A Session-to-session Transfer Learning Approach for Non-stationary EEG. In Ishibuchi H, Kwoh CK, Tan AH, Srinivasan D, Miao C, Trivedi A, Crockett K, editors, Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022. Institute of Electrical and Electronics Engineers Inc. 2022. p. 164-170. (Proceedings of the 2022 IEEE Symposium Series on Computational Intelligence, SSCI 2022). doi: 10.1109/SSCI51031.2022.10022274

Online Adaptive CNN: A Session-to-session Transfer Learning Approach for Non-stationary EEG

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this