Epileptic seizure detection in EEG signals using sparse multiscale radial basis function networks and the Fisher vector approach

Yang Li, Wei Gang Cui (Corresponding author), Hui Huang, Yu Zhu Guo, Ke Li, Tao Tan

Research output: Contribution to journalArticleAcademicpeer-review

7 Citations (Scopus)

Abstract

Detecting epileptic seizures in electroencephalography (EEG) signals is a challenging task due to nonstationary processes of brain activities. Currently, the epilepsy is mainly detected by clinicians based on visual observation of EEG recordings, which is generally time consuming and sensitive to bias. This paper presents a novel automatic seizure detection method based on the multiscale radial basis function (MRBF) networks and the Fisher vector (FV) encoding. Specifically, the MRBF networks are first used to obtain high-resolution time-frequency (TF) images for feature extraction, where both a modified particle swarm optimization (MPSO) method and an orthogonal least squares (OLS) algorithm are implemented to determine optimal scales and detect a parsimonious model structure. Gray level co-occurrence matrix (GLCM) texture descriptors and the FV, which contribute to high-dimensional vectors, are then adopted to achieve discriminative features based on five frequency subbands of clinical interests from TF images. Furthermore, the dimensionality of the original feature space can be effectively reduced by the t-test statistical tool before feeding compact features into the SVM classifier for seizure detection. Finally, the classification performance of the proposed method is evaluated by using two widely used EEG database, and is observed to provide good classification accuracy on both datasets. Experimental results demonstrate that our proposed method is a powerful tool in detecting epileptic seizures.

Original languageEnglish
Pages (from-to)96-106
Number of pages11
JournalKnowledge-Based Systems
Volume164
DOIs
Publication statusPublished - 15 Jan 2019

Fingerprint

Radial basis function networks
Electroencephalography
Statistical tests
Model structures
Particle swarm optimization (PSO)
Feature extraction
Brain
Classifiers
Textures
Radial basis function

Keywords

  • Electroencephalography (EEG)
  • Fisher vector
  • Modified particle swarm optimization (MPSO)
  • Multiscale radial basis functions (MRBF)
  • Orthogonal least squares (OLS)
  • Seizure detection

Cite this

@article{885a5b7b910744df96b81111c383387d,
title = "Epileptic seizure detection in EEG signals using sparse multiscale radial basis function networks and the Fisher vector approach",
abstract = "Detecting epileptic seizures in electroencephalography (EEG) signals is a challenging task due to nonstationary processes of brain activities. Currently, the epilepsy is mainly detected by clinicians based on visual observation of EEG recordings, which is generally time consuming and sensitive to bias. This paper presents a novel automatic seizure detection method based on the multiscale radial basis function (MRBF) networks and the Fisher vector (FV) encoding. Specifically, the MRBF networks are first used to obtain high-resolution time-frequency (TF) images for feature extraction, where both a modified particle swarm optimization (MPSO) method and an orthogonal least squares (OLS) algorithm are implemented to determine optimal scales and detect a parsimonious model structure. Gray level co-occurrence matrix (GLCM) texture descriptors and the FV, which contribute to high-dimensional vectors, are then adopted to achieve discriminative features based on five frequency subbands of clinical interests from TF images. Furthermore, the dimensionality of the original feature space can be effectively reduced by the t-test statistical tool before feeding compact features into the SVM classifier for seizure detection. Finally, the classification performance of the proposed method is evaluated by using two widely used EEG database, and is observed to provide good classification accuracy on both datasets. Experimental results demonstrate that our proposed method is a powerful tool in detecting epileptic seizures.",
keywords = "Electroencephalography (EEG), Fisher vector, Modified particle swarm optimization (MPSO), Multiscale radial basis functions (MRBF), Orthogonal least squares (OLS), Seizure detection",
author = "Yang Li and Cui, {Wei Gang} and Hui Huang and Guo, {Yu Zhu} and Ke Li and Tao Tan",
year = "2019",
month = "1",
day = "15",
doi = "10.1016/j.knosys.2018.10.029",
language = "English",
volume = "164",
pages = "96--106",
journal = "Knowledge-Based Systems",
issn = "0950-7051",
publisher = "Elsevier",

}

Epileptic seizure detection in EEG signals using sparse multiscale radial basis function networks and the Fisher vector approach. / Li, Yang; Cui, Wei Gang (Corresponding author); Huang, Hui; Guo, Yu Zhu; Li, Ke; Tan, Tao.

In: Knowledge-Based Systems, Vol. 164, 15.01.2019, p. 96-106.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Epileptic seizure detection in EEG signals using sparse multiscale radial basis function networks and the Fisher vector approach

AU - Li, Yang

AU - Cui, Wei Gang

AU - Huang, Hui

AU - Guo, Yu Zhu

AU - Li, Ke

AU - Tan, Tao

PY - 2019/1/15

Y1 - 2019/1/15

N2 - Detecting epileptic seizures in electroencephalography (EEG) signals is a challenging task due to nonstationary processes of brain activities. Currently, the epilepsy is mainly detected by clinicians based on visual observation of EEG recordings, which is generally time consuming and sensitive to bias. This paper presents a novel automatic seizure detection method based on the multiscale radial basis function (MRBF) networks and the Fisher vector (FV) encoding. Specifically, the MRBF networks are first used to obtain high-resolution time-frequency (TF) images for feature extraction, where both a modified particle swarm optimization (MPSO) method and an orthogonal least squares (OLS) algorithm are implemented to determine optimal scales and detect a parsimonious model structure. Gray level co-occurrence matrix (GLCM) texture descriptors and the FV, which contribute to high-dimensional vectors, are then adopted to achieve discriminative features based on five frequency subbands of clinical interests from TF images. Furthermore, the dimensionality of the original feature space can be effectively reduced by the t-test statistical tool before feeding compact features into the SVM classifier for seizure detection. Finally, the classification performance of the proposed method is evaluated by using two widely used EEG database, and is observed to provide good classification accuracy on both datasets. Experimental results demonstrate that our proposed method is a powerful tool in detecting epileptic seizures.

AB - Detecting epileptic seizures in electroencephalography (EEG) signals is a challenging task due to nonstationary processes of brain activities. Currently, the epilepsy is mainly detected by clinicians based on visual observation of EEG recordings, which is generally time consuming and sensitive to bias. This paper presents a novel automatic seizure detection method based on the multiscale radial basis function (MRBF) networks and the Fisher vector (FV) encoding. Specifically, the MRBF networks are first used to obtain high-resolution time-frequency (TF) images for feature extraction, where both a modified particle swarm optimization (MPSO) method and an orthogonal least squares (OLS) algorithm are implemented to determine optimal scales and detect a parsimonious model structure. Gray level co-occurrence matrix (GLCM) texture descriptors and the FV, which contribute to high-dimensional vectors, are then adopted to achieve discriminative features based on five frequency subbands of clinical interests from TF images. Furthermore, the dimensionality of the original feature space can be effectively reduced by the t-test statistical tool before feeding compact features into the SVM classifier for seizure detection. Finally, the classification performance of the proposed method is evaluated by using two widely used EEG database, and is observed to provide good classification accuracy on both datasets. Experimental results demonstrate that our proposed method is a powerful tool in detecting epileptic seizures.

KW - Electroencephalography (EEG)

KW - Fisher vector

KW - Modified particle swarm optimization (MPSO)

KW - Multiscale radial basis functions (MRBF)

KW - Orthogonal least squares (OLS)

KW - Seizure detection

UR - http://www.scopus.com/inward/record.url?scp=85055887576&partnerID=8YFLogxK

U2 - 10.1016/j.knosys.2018.10.029

DO - 10.1016/j.knosys.2018.10.029

M3 - Article

AN - SCOPUS:85055887576

VL - 164

SP - 96

EP - 106

JO - Knowledge-Based Systems

JF - Knowledge-Based Systems

SN - 0950-7051

ER -