Spatio-spectral representation learning for electroencephalographic gait-pattern classification

Sim Kuan Goh, Hussein A. Abbass, Kay Chen Tan, Mohamed Almamun, Nitish Thakor, Anastasios Bezerianos, Junhua Li

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

The brain plays a pivotal role in locomotion by coordinating muscles through interconnections that get established by the peripheral nervous system. To date, many attempts have been made to reveal the underlying mechanisms of humans' gait. However, decoding cortical processes associated with different walking conditions using EEG signals for gait-pattern classification is a less-explored research area. In this paper, we design an EEG-based experiment with four walking conditions (i.e., free walking, and exoskeleton-assisted walking at zero, low, and high assistive forces by the use of a unilateral exoskeleton to right lower limb). We proposed spatio-spectral representation learning (SSRL), a deep neural network topology with shared weights to learn the spatial and spectral representations of multi-channel EEG signals during walking. Adoption of weight sharing reduces the number of free parameters, while learning spatial and spectral equivariant features. SSRL outperformed state-of-the-art methods in decoding gait patterns, achieving a classification accuracy of 77.8%. Moreover, the features extracted in the intermediate layer of SSRL were observed to be more discriminative than the hand-crafted features. When analyzing the weights of the proposed model, we found an intriguing spatial distribution that is consistent with the distribution found in well-known motor-activated cortical regions. Our results show that SSRL advances the ability to decode human locomotion and it could have important implications for exoskeleton design, rehabilitation processes, and clinical diagnosis.

Original languageEnglish
Article number8428659
Pages (from-to)1858-1867
Number of pages10
JournalIEEE Transactions on Neural Systems and Rehabilitation Engineering
Volume26
Issue number9
DOIs
Publication statusPublished - Sep 1 2018

Fingerprint

Electroencephalography
Gait
Walking
Pattern recognition
Learning
Decoding
Locomotion
Weights and Measures
Neurology
Patient rehabilitation
Spatial distribution
Muscle
Brain
Topology
Aptitude
Peripheral Nervous System
Lower Extremity
Rehabilitation
Hand
Muscles

Keywords

  • convolutional neural network
  • electroencephalogram (EEG)
  • exoskeleton
  • gait pattern
  • Spatio-spectral representation learning

ASJC Scopus subject areas

  • Neuroscience(all)
  • Biomedical Engineering
  • Computer Science Applications

Cite this

Spatio-spectral representation learning for electroencephalographic gait-pattern classification. / Goh, Sim Kuan; Abbass, Hussein A.; Tan, Kay Chen; Almamun, Mohamed; Thakor, Nitish; Bezerianos, Anastasios; Li, Junhua.

In: IEEE Transactions on Neural Systems and Rehabilitation Engineering, Vol. 26, No. 9, 8428659, 01.09.2018, p. 1858-1867.

Research output: Contribution to journalArticle

Goh, Sim Kuan ; Abbass, Hussein A. ; Tan, Kay Chen ; Almamun, Mohamed ; Thakor, Nitish ; Bezerianos, Anastasios ; Li, Junhua. / Spatio-spectral representation learning for electroencephalographic gait-pattern classification. In: IEEE Transactions on Neural Systems and Rehabilitation Engineering. 2018 ; Vol. 26, No. 9. pp. 1858-1867.
@article{7a71b2ffedf94f6796065419e8c6e819,
title = "Spatio-spectral representation learning for electroencephalographic gait-pattern classification",
abstract = "The brain plays a pivotal role in locomotion by coordinating muscles through interconnections that get established by the peripheral nervous system. To date, many attempts have been made to reveal the underlying mechanisms of humans' gait. However, decoding cortical processes associated with different walking conditions using EEG signals for gait-pattern classification is a less-explored research area. In this paper, we design an EEG-based experiment with four walking conditions (i.e., free walking, and exoskeleton-assisted walking at zero, low, and high assistive forces by the use of a unilateral exoskeleton to right lower limb). We proposed spatio-spectral representation learning (SSRL), a deep neural network topology with shared weights to learn the spatial and spectral representations of multi-channel EEG signals during walking. Adoption of weight sharing reduces the number of free parameters, while learning spatial and spectral equivariant features. SSRL outperformed state-of-the-art methods in decoding gait patterns, achieving a classification accuracy of 77.8{\%}. Moreover, the features extracted in the intermediate layer of SSRL were observed to be more discriminative than the hand-crafted features. When analyzing the weights of the proposed model, we found an intriguing spatial distribution that is consistent with the distribution found in well-known motor-activated cortical regions. Our results show that SSRL advances the ability to decode human locomotion and it could have important implications for exoskeleton design, rehabilitation processes, and clinical diagnosis.",
keywords = "convolutional neural network, electroencephalogram (EEG), exoskeleton, gait pattern, Spatio-spectral representation learning",
author = "Goh, {Sim Kuan} and Abbass, {Hussein A.} and Tan, {Kay Chen} and Mohamed Almamun and Nitish Thakor and Anastasios Bezerianos and Junhua Li",
year = "2018",
month = "9",
day = "1",
doi = "10.1109/TNSRE.2018.2864119",
language = "English",
volume = "26",
pages = "1858--1867",
journal = "IEEE Transactions on Neural Systems and Rehabilitation Engineering",
issn = "1534-4320",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "9",

}

TY - JOUR

T1 - Spatio-spectral representation learning for electroencephalographic gait-pattern classification

AU - Goh, Sim Kuan

AU - Abbass, Hussein A.

AU - Tan, Kay Chen

AU - Almamun, Mohamed

AU - Thakor, Nitish

AU - Bezerianos, Anastasios

AU - Li, Junhua

PY - 2018/9/1

Y1 - 2018/9/1

N2 - The brain plays a pivotal role in locomotion by coordinating muscles through interconnections that get established by the peripheral nervous system. To date, many attempts have been made to reveal the underlying mechanisms of humans' gait. However, decoding cortical processes associated with different walking conditions using EEG signals for gait-pattern classification is a less-explored research area. In this paper, we design an EEG-based experiment with four walking conditions (i.e., free walking, and exoskeleton-assisted walking at zero, low, and high assistive forces by the use of a unilateral exoskeleton to right lower limb). We proposed spatio-spectral representation learning (SSRL), a deep neural network topology with shared weights to learn the spatial and spectral representations of multi-channel EEG signals during walking. Adoption of weight sharing reduces the number of free parameters, while learning spatial and spectral equivariant features. SSRL outperformed state-of-the-art methods in decoding gait patterns, achieving a classification accuracy of 77.8%. Moreover, the features extracted in the intermediate layer of SSRL were observed to be more discriminative than the hand-crafted features. When analyzing the weights of the proposed model, we found an intriguing spatial distribution that is consistent with the distribution found in well-known motor-activated cortical regions. Our results show that SSRL advances the ability to decode human locomotion and it could have important implications for exoskeleton design, rehabilitation processes, and clinical diagnosis.

AB - The brain plays a pivotal role in locomotion by coordinating muscles through interconnections that get established by the peripheral nervous system. To date, many attempts have been made to reveal the underlying mechanisms of humans' gait. However, decoding cortical processes associated with different walking conditions using EEG signals for gait-pattern classification is a less-explored research area. In this paper, we design an EEG-based experiment with four walking conditions (i.e., free walking, and exoskeleton-assisted walking at zero, low, and high assistive forces by the use of a unilateral exoskeleton to right lower limb). We proposed spatio-spectral representation learning (SSRL), a deep neural network topology with shared weights to learn the spatial and spectral representations of multi-channel EEG signals during walking. Adoption of weight sharing reduces the number of free parameters, while learning spatial and spectral equivariant features. SSRL outperformed state-of-the-art methods in decoding gait patterns, achieving a classification accuracy of 77.8%. Moreover, the features extracted in the intermediate layer of SSRL were observed to be more discriminative than the hand-crafted features. When analyzing the weights of the proposed model, we found an intriguing spatial distribution that is consistent with the distribution found in well-known motor-activated cortical regions. Our results show that SSRL advances the ability to decode human locomotion and it could have important implications for exoskeleton design, rehabilitation processes, and clinical diagnosis.

KW - convolutional neural network

KW - electroencephalogram (EEG)

KW - exoskeleton

KW - gait pattern

KW - Spatio-spectral representation learning

UR - http://www.scopus.com/inward/record.url?scp=85051412275&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85051412275&partnerID=8YFLogxK

U2 - 10.1109/TNSRE.2018.2864119

DO - 10.1109/TNSRE.2018.2864119

M3 - Article

VL - 26

SP - 1858

EP - 1867

JO - IEEE Transactions on Neural Systems and Rehabilitation Engineering

JF - IEEE Transactions on Neural Systems and Rehabilitation Engineering

SN - 1534-4320

IS - 9

M1 - 8428659

ER -