Novel approaches of EEG signal classification using IMF bandwidth and DCT frequency

Mohammad Parvez, Manoranjan Paul

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Electroencephalogram (EEG) is a record of ongoing electrical signal to represent the human brain activity. It has great potential for the diagnosis to treatment of mental disorder and brain diseases such as epileptic seizure. Features extraction and classi¯cation is a crucial task to detect the stage of ictal (i.e. seizure period) and interictal (i.e. periodbetween seizures) EEG signals for the treatment and precaution of the patient. However, existing seizure and nonseizure feature extraction techniques are not good enough for the classi¯cation of ictal and interictal EEG signals considering their non-abrupt phenomena and inconsistency in di®erent brain locations. In this paper, we present newapproaches for feature extraction using high-frequency components from discrete cosine transformation (DCT) and intrinsic mode function (IMF) extracted from empirical mode decomposition (EMD). These features are then used as an input to least square-support vector machine (LV-SVM) to classify ictal and interictal EEG signals. Experimental results show that the proposed methods outperform the existing state-of-the-art method for better classi¯cation in terms of sensitivity, speci¯city, and accuracy with greater consistence of ictal and interictal period of epilepsy for benchmark dataset from di®erent brain locations.
Original languageEnglish
Pages (from-to)1-9
Number of pages9
JournalBiomedical Engineering - Applications, Basis and Communications
Volume27
Issue number3
DOIs
Publication statusPublished - Jun 2015

Fingerprint

Electroencephalography
Brain
Stroke
Bandwidth
Feature extraction
Seizures
Epilepsy
Benchmarking
Brain Diseases
Least-Squares Analysis
Mental Disorders
Human Activities
Support vector machines
Decomposition
Therapeutics

Cite this

@article{1c6c839c2e174e23a00b3318dc06be01,
title = "Novel approaches of EEG signal classification using IMF bandwidth and DCT frequency",
abstract = "Electroencephalogram (EEG) is a record of ongoing electrical signal to represent the human brain activity. It has great potential for the diagnosis to treatment of mental disorder and brain diseases such as epileptic seizure. Features extraction and classi{\~A}‚{\^A}¯cation is a crucial task to detect the stage of ictal (i.e. seizure period) and interictal (i.e. periodbetween seizures) EEG signals for the treatment and precaution of the patient. However, existing seizure and nonseizure feature extraction techniques are not good enough for the classi{\~A}‚{\^A}¯cation of ictal and interictal EEG signals considering their non-abrupt phenomena and inconsistency in di{\~A}‚{\^A}{\circledR}erent brain locations. In this paper, we present newapproaches for feature extraction using high-frequency components from discrete cosine transformation (DCT) and intrinsic mode function (IMF) extracted from empirical mode decomposition (EMD). These features are then used as an input to least square-support vector machine (LV-SVM) to classify ictal and interictal EEG signals. Experimental results show that the proposed methods outperform the existing state-of-the-art method for better classi{\~A}‚{\^A}¯cation in terms of sensitivity, speci{\~A}‚{\^A}¯city, and accuracy with greater consistence of ictal and interictal period of epilepsy for benchmark dataset from di{\~A}‚{\^A}{\circledR}erent brain locations.",
keywords = "EEG, EMD, Epilepsy, IMF, Ictal, Interictal, LS-SVM, Seizure",
author = "Mohammad Parvez and Manoranjan Paul",
note = "Imported on 12 Apr 2017 - DigiTool details were: month (773h) = June; Journal title (773t) = Biomedical Engineering: Applications, Basis and Communications. ISSNs: 1016-2372;",
year = "2015",
month = "6",
doi = "10.4015/S1016237215500271",
language = "English",
volume = "27",
pages = "1--9",
journal = "Biomedical Engineering - Applications, Basis and Communications",
issn = "1016-2372",
publisher = "World Scientific Publishing",
number = "3",

}

Novel approaches of EEG signal classification using IMF bandwidth and DCT frequency. / Parvez, Mohammad; Paul, Manoranjan.

In: Biomedical Engineering - Applications, Basis and Communications, Vol. 27, No. 3, 06.2015, p. 1-9.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Novel approaches of EEG signal classification using IMF bandwidth and DCT frequency

AU - Parvez, Mohammad

AU - Paul, Manoranjan

N1 - Imported on 12 Apr 2017 - DigiTool details were: month (773h) = June; Journal title (773t) = Biomedical Engineering: Applications, Basis and Communications. ISSNs: 1016-2372;

PY - 2015/6

Y1 - 2015/6

N2 - Electroencephalogram (EEG) is a record of ongoing electrical signal to represent the human brain activity. It has great potential for the diagnosis to treatment of mental disorder and brain diseases such as epileptic seizure. Features extraction and classi¯cation is a crucial task to detect the stage of ictal (i.e. seizure period) and interictal (i.e. periodbetween seizures) EEG signals for the treatment and precaution of the patient. However, existing seizure and nonseizure feature extraction techniques are not good enough for the classi¯cation of ictal and interictal EEG signals considering their non-abrupt phenomena and inconsistency in di®erent brain locations. In this paper, we present newapproaches for feature extraction using high-frequency components from discrete cosine transformation (DCT) and intrinsic mode function (IMF) extracted from empirical mode decomposition (EMD). These features are then used as an input to least square-support vector machine (LV-SVM) to classify ictal and interictal EEG signals. Experimental results show that the proposed methods outperform the existing state-of-the-art method for better classi¯cation in terms of sensitivity, speci¯city, and accuracy with greater consistence of ictal and interictal period of epilepsy for benchmark dataset from di®erent brain locations.

AB - Electroencephalogram (EEG) is a record of ongoing electrical signal to represent the human brain activity. It has great potential for the diagnosis to treatment of mental disorder and brain diseases such as epileptic seizure. Features extraction and classi¯cation is a crucial task to detect the stage of ictal (i.e. seizure period) and interictal (i.e. periodbetween seizures) EEG signals for the treatment and precaution of the patient. However, existing seizure and nonseizure feature extraction techniques are not good enough for the classi¯cation of ictal and interictal EEG signals considering their non-abrupt phenomena and inconsistency in di®erent brain locations. In this paper, we present newapproaches for feature extraction using high-frequency components from discrete cosine transformation (DCT) and intrinsic mode function (IMF) extracted from empirical mode decomposition (EMD). These features are then used as an input to least square-support vector machine (LV-SVM) to classify ictal and interictal EEG signals. Experimental results show that the proposed methods outperform the existing state-of-the-art method for better classi¯cation in terms of sensitivity, speci¯city, and accuracy with greater consistence of ictal and interictal period of epilepsy for benchmark dataset from di®erent brain locations.

KW - EEG

KW - EMD

KW - Epilepsy

KW - IMF

KW - Ictal

KW - Interictal

KW - LS-SVM

KW - Seizure

U2 - 10.4015/S1016237215500271

DO - 10.4015/S1016237215500271

M3 - Article

VL - 27

SP - 1

EP - 9

JO - Biomedical Engineering - Applications, Basis and Communications

JF - Biomedical Engineering - Applications, Basis and Communications

SN - 1016-2372

IS - 3

ER -