System derived spatial-temporal CNN for high-density fNIRS BCI

Robin Dale; Thomas D. O'Sullivan; Scott Howard; Felipe Orihuela-Espina; Hamid Dehghani

doi:10.1109/OJEMB.2023.3248492

System derived spatial-temporal CNN for high-density fNIRS BCI

Robin Dale, Thomas D. O'Sullivan, Scott Howard, Felipe Orihuela-Espina, Hamid Dehghani

Computer Science

Research output: Contribution to journal › Article › peer-review

14 Downloads (Pure)

Abstract

An intuitive and generalisable approach to spatial-temporal feature extraction for high-density (HD) functional Near-Infrared Spectroscopy (fNIRS) brain-computer interface (BCI) is proposed, demonstrated here using Frequency-Domain (FD) fNIRS for motor-task classification. Enabled by the HD probe design, layered topographical maps of Oxy/deOxy Haemoglobin changes are used to train a 3D convolutional neural network (CNN), enabling simultaneous extraction of spatial and temporal features. The proposed spatial-temporal CNN is shown to effectively exploit the spatial relationships in HD fNIRS measurements to improve the classification of the functional haemodynamic response, achieving an average F1 score of 0.69 across seven subjects in a mixed subjects training scheme, and improving subject-independent classification as compared to a standard temporal CNN.

Original language	English
Article number	10073629
Pages (from-to)	85-95
Number of pages	11
Journal	IEEE Open Journal of Engineering in Medicine and Biology
Volume	4
Early online date	16 Mar 2023
DOIs	https://doi.org/10.1109/OJEMB.2023.3248492
Publication status	Published - 22 May 2023

Bibliographical note

Funding Information:
This work was supported by the National Institute of Biomedical Imaging, and Bioengineering (NIBIB) of the National Institutes of Health (NIH) under Award R01EB029595.

Publisher Copyright:
© 2020 IEEE.

Keywords

Functional near-infrared spectroscopy
Feature extraction
Probes
Convolutional neural networks
Task analysis
Hardware
Biological neural networks

ASJC Scopus subject areas

Biomedical Engineering

Access to Document

10.1109/OJEMB.2023.3248492Licence: Creative Commons: Attribution (CC BY)

DaleR2023SystemFinal published version, 3.65 MBLicence: Creative Commons: Attribution (CC BY)

https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10073629Licence: Creative Commons: Attribution (CC BY)

Cite this

@article{16bda0bb874d4a819f661a66b1a7d361,

title = "System derived spatial-temporal CNN for high-density fNIRS BCI",

abstract = "An intuitive and generalisable approach to spatial-temporal feature extraction for high-density (HD) functional Near-Infrared Spectroscopy (fNIRS) brain-computer interface (BCI) is proposed, demonstrated here using Frequency-Domain (FD) fNIRS for motor-task classification. Enabled by the HD probe design, layered topographical maps of Oxy/deOxy Haemoglobin changes are used to train a 3D convolutional neural network (CNN), enabling simultaneous extraction of spatial and temporal features. The proposed spatial-temporal CNN is shown to effectively exploit the spatial relationships in HD fNIRS measurements to improve the classification of the functional haemodynamic response, achieving an average F1 score of 0.69 across seven subjects in a mixed subjects training scheme, and improving subject-independent classification as compared to a standard temporal CNN.",

keywords = "Functional near-infrared spectroscopy, Feature extraction, Probes, Convolutional neural networks, Task analysis, Hardware, Biological neural networks",

author = "Robin Dale and O'Sullivan, {Thomas D.} and Scott Howard and Felipe Orihuela-Espina and Hamid Dehghani",

note = "Funding Information: This work was supported by the National Institute of Biomedical Imaging, and Bioengineering (NIBIB) of the National Institutes of Health (NIH) under Award R01EB029595. Publisher Copyright: {\textcopyright} 2020 IEEE.",

year = "2023",

month = may,

day = "22",

doi = "10.1109/OJEMB.2023.3248492",

language = "English",

volume = "4",

pages = "85--95",

journal = "IEEE Open Journal of Engineering in Medicine and Biology",

issn = "2644-1276",

publisher = "IEEE",

}

TY - JOUR

T1 - System derived spatial-temporal CNN for high-density fNIRS BCI

AU - Dale, Robin

AU - O'Sullivan, Thomas D.

AU - Howard, Scott

AU - Orihuela-Espina, Felipe

AU - Dehghani, Hamid

N1 - Funding Information: This work was supported by the National Institute of Biomedical Imaging, and Bioengineering (NIBIB) of the National Institutes of Health (NIH) under Award R01EB029595. Publisher Copyright: © 2020 IEEE.

PY - 2023/5/22

Y1 - 2023/5/22

N2 - An intuitive and generalisable approach to spatial-temporal feature extraction for high-density (HD) functional Near-Infrared Spectroscopy (fNIRS) brain-computer interface (BCI) is proposed, demonstrated here using Frequency-Domain (FD) fNIRS for motor-task classification. Enabled by the HD probe design, layered topographical maps of Oxy/deOxy Haemoglobin changes are used to train a 3D convolutional neural network (CNN), enabling simultaneous extraction of spatial and temporal features. The proposed spatial-temporal CNN is shown to effectively exploit the spatial relationships in HD fNIRS measurements to improve the classification of the functional haemodynamic response, achieving an average F1 score of 0.69 across seven subjects in a mixed subjects training scheme, and improving subject-independent classification as compared to a standard temporal CNN.

AB - An intuitive and generalisable approach to spatial-temporal feature extraction for high-density (HD) functional Near-Infrared Spectroscopy (fNIRS) brain-computer interface (BCI) is proposed, demonstrated here using Frequency-Domain (FD) fNIRS for motor-task classification. Enabled by the HD probe design, layered topographical maps of Oxy/deOxy Haemoglobin changes are used to train a 3D convolutional neural network (CNN), enabling simultaneous extraction of spatial and temporal features. The proposed spatial-temporal CNN is shown to effectively exploit the spatial relationships in HD fNIRS measurements to improve the classification of the functional haemodynamic response, achieving an average F1 score of 0.69 across seven subjects in a mixed subjects training scheme, and improving subject-independent classification as compared to a standard temporal CNN.

KW - Functional near-infrared spectroscopy

KW - Feature extraction

KW - Probes

KW - Convolutional neural networks

KW - Task analysis

KW - Hardware

KW - Biological neural networks

UR - https://ieeexplore.ieee.org/document/10073629/

UR - http://www.scopus.com/inward/record.url?scp=85151561259&partnerID=8YFLogxK

U2 - 10.1109/OJEMB.2023.3248492

DO - 10.1109/OJEMB.2023.3248492

M3 - Article

SN - 2644-1276

VL - 4

SP - 85

EP - 95

JO - IEEE Open Journal of Engineering in Medicine and Biology

JF - IEEE Open Journal of Engineering in Medicine and Biology

M1 - 10073629

ER -

System derived spatial-temporal CNN for high-density fNIRS BCI

Abstract

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Fingerprint

Cite this