Applying deep-learning to a top-down SSVEP BMI

Min Hee Ahn; Byoung-Kyong Min

doi:10.1109/IWW-BCI.2018.8311526

Applying deep-learning to a top-down SSVEP BMI

Min Hee Ahn, Byoung-Kyong Min

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

5 Citations (Scopus)

Abstract

Brain-machine interfaces (BMIs) enable humans to control devices by modulating their brain signals. As the current BMI technology has several obstacles to overcome, additional sources of brain activity need to be explored. It seems plausible that the brain activity associated with top-down cognitive functions could open a new prospect in the field of BMIs. As top-down cognitive BMIs could exploit neural signals from more diverse networks, a deep-learning approach with complex hidden layers may provide a more optimal decoding performance. In this study, using our top-down steady-state visual evoked potential (SSVEP) paradigm (N = 20), we observed that the decoding accuracy (48.42%) of a deep-learning algorithm with a sigmoid activation function was significantly higher than that of regularized linear discriminant analysis (rLDA) with shrinkage (42.52%; t(19) =-3.183, p < 0.01), used in our previous study. Therefore, a deep-learning approach seems to be more optimized for classification in the top-down cognitive BMI paradigm.

Original language	English
Title of host publication	2018 6th International Conference on Brain-Computer Interface, BCI 2018
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	1-3
Number of pages	3
ISBN (Electronic)	9781538625743
DOIs	https://doi.org/10.1109/IWW-BCI.2018.8311526
Publication status	Published - 2018 Mar 9
Event	6th International Conference on Brain-Computer Interface, BCI 2018 - GangWon, Korea, Republic of Duration: 2018 Jan 15 → 2018 Jan 17

Publication series

Name	2018 6th International Conference on Brain-Computer Interface, BCI 2018
Volume	2018-January

Other

Other	6th International Conference on Brain-Computer Interface, BCI 2018
Country/Territory	Korea, Republic of
City	GangWon
Period	18/1/15 → 18/1/17

Bibliographical note

Funding Information:
This work was supported by the Basic Science Research program (grant numbers 2015R1A1A1A05027233), the ICT R&D program of MSIP/Institute for Information & Communications Technology Promotion (IITP; grant number 2017-0-00432), and the Information Technology Research Center (ITRC) support program (grant number IITP-2017-2016-0-00464) supervised by the IITP, which are funded by the Korean government (MSIT) through the National Research Foundation of Korea. The authors declare that they have no competing interests.

Publisher Copyright:
© 2018 IEEE.

Keywords

BMI
DNN
EEG
SSVEP
deep-learning
top-down

ASJC Scopus subject areas

Artificial Intelligence
Human-Computer Interaction
Behavioral Neuroscience

Access to Document

10.1109/IWW-BCI.2018.8311526

Cite this

Applying deep-learning to a top-down SSVEP BMI. / Ahn, Min Hee; Min, Byoung-Kyong.
2018 6th International Conference on Brain-Computer Interface, BCI 2018. Institute of Electrical and Electronics Engineers Inc., 2018. p. 1-3 (2018 6th International Conference on Brain-Computer Interface, BCI 2018; Vol. 2018-January).

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

Ahn, MH & Min, B-K 2018, Applying deep-learning to a top-down SSVEP BMI. in 2018 6th International Conference on Brain-Computer Interface, BCI 2018. 2018 6th International Conference on Brain-Computer Interface, BCI 2018, vol. 2018-January, Institute of Electrical and Electronics Engineers Inc., pp. 1-3, 6th International Conference on Brain-Computer Interface, BCI 2018, GangWon, Korea, Republic of, 18/1/15. https://doi.org/10.1109/IWW-BCI.2018.8311526

@inproceedings{011a6ae3ff8a44eb90123a40683e18d2,

title = "Applying deep-learning to a top-down SSVEP BMI",

abstract = "Brain-machine interfaces (BMIs) enable humans to control devices by modulating their brain signals. As the current BMI technology has several obstacles to overcome, additional sources of brain activity need to be explored. It seems plausible that the brain activity associated with top-down cognitive functions could open a new prospect in the field of BMIs. As top-down cognitive BMIs could exploit neural signals from more diverse networks, a deep-learning approach with complex hidden layers may provide a more optimal decoding performance. In this study, using our top-down steady-state visual evoked potential (SSVEP) paradigm (N = 20), we observed that the decoding accuracy (48.42%) of a deep-learning algorithm with a sigmoid activation function was significantly higher than that of regularized linear discriminant analysis (rLDA) with shrinkage (42.52%; t(19) =-3.183, p < 0.01), used in our previous study. Therefore, a deep-learning approach seems to be more optimized for classification in the top-down cognitive BMI paradigm.",

keywords = "BMI, DNN, EEG, SSVEP, deep-learning, top-down",

author = "Ahn, {Min Hee} and Byoung-Kyong Min",

note = "Funding Information: This work was supported by the Basic Science Research program (grant numbers 2015R1A1A1A05027233), the ICT R&D program of MSIP/Institute for Information & Communications Technology Promotion (IITP; grant number 2017-0-00432), and the Information Technology Research Center (ITRC) support program (grant number IITP-2017-2016-0-00464) supervised by the IITP, which are funded by the Korean government (MSIT) through the National Research Foundation of Korea. The authors declare that they have no competing interests. Publisher Copyright: {\textcopyright} 2018 IEEE.; 6th International Conference on Brain-Computer Interface, BCI 2018 ; Conference date: 15-01-2018 Through 17-01-2018",

year = "2018",

month = mar,

day = "9",

doi = "10.1109/IWW-BCI.2018.8311526",

language = "English",

series = "2018 6th International Conference on Brain-Computer Interface, BCI 2018",

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

pages = "1--3",

booktitle = "2018 6th International Conference on Brain-Computer Interface, BCI 2018",

}

TY - GEN

T1 - Applying deep-learning to a top-down SSVEP BMI

AU - Ahn, Min Hee

AU - Min, Byoung-Kyong

N1 - Funding Information: This work was supported by the Basic Science Research program (grant numbers 2015R1A1A1A05027233), the ICT R&D program of MSIP/Institute for Information & Communications Technology Promotion (IITP; grant number 2017-0-00432), and the Information Technology Research Center (ITRC) support program (grant number IITP-2017-2016-0-00464) supervised by the IITP, which are funded by the Korean government (MSIT) through the National Research Foundation of Korea. The authors declare that they have no competing interests. Publisher Copyright: © 2018 IEEE.

PY - 2018/3/9

Y1 - 2018/3/9

N2 - Brain-machine interfaces (BMIs) enable humans to control devices by modulating their brain signals. As the current BMI technology has several obstacles to overcome, additional sources of brain activity need to be explored. It seems plausible that the brain activity associated with top-down cognitive functions could open a new prospect in the field of BMIs. As top-down cognitive BMIs could exploit neural signals from more diverse networks, a deep-learning approach with complex hidden layers may provide a more optimal decoding performance. In this study, using our top-down steady-state visual evoked potential (SSVEP) paradigm (N = 20), we observed that the decoding accuracy (48.42%) of a deep-learning algorithm with a sigmoid activation function was significantly higher than that of regularized linear discriminant analysis (rLDA) with shrinkage (42.52%; t(19) =-3.183, p < 0.01), used in our previous study. Therefore, a deep-learning approach seems to be more optimized for classification in the top-down cognitive BMI paradigm.

AB - Brain-machine interfaces (BMIs) enable humans to control devices by modulating their brain signals. As the current BMI technology has several obstacles to overcome, additional sources of brain activity need to be explored. It seems plausible that the brain activity associated with top-down cognitive functions could open a new prospect in the field of BMIs. As top-down cognitive BMIs could exploit neural signals from more diverse networks, a deep-learning approach with complex hidden layers may provide a more optimal decoding performance. In this study, using our top-down steady-state visual evoked potential (SSVEP) paradigm (N = 20), we observed that the decoding accuracy (48.42%) of a deep-learning algorithm with a sigmoid activation function was significantly higher than that of regularized linear discriminant analysis (rLDA) with shrinkage (42.52%; t(19) =-3.183, p < 0.01), used in our previous study. Therefore, a deep-learning approach seems to be more optimized for classification in the top-down cognitive BMI paradigm.

KW - BMI

KW - DNN

KW - EEG

KW - SSVEP

KW - deep-learning

KW - top-down

UR - http://www.scopus.com/inward/record.url?scp=85050809368&partnerID=8YFLogxK

U2 - 10.1109/IWW-BCI.2018.8311526

DO - 10.1109/IWW-BCI.2018.8311526

M3 - Conference contribution

AN - SCOPUS:85050809368

T3 - 2018 6th International Conference on Brain-Computer Interface, BCI 2018

SP - 1

EP - 3

BT - 2018 6th International Conference on Brain-Computer Interface, BCI 2018

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 6th International Conference on Brain-Computer Interface, BCI 2018

Y2 - 15 January 2018 through 17 January 2018

ER -

Applying deep-learning to a top-down SSVEP BMI

Abstract

Publication series

Other

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this