Seizure Prediction using Undulated Global and Local Features

Mohammad Zavid Parvez, Manoranjan Paul

Research output: Contribution to journalArticle

15 Citations (Scopus)
7 Downloads (Pure)

Abstract

In this study, a seizure prediction method is proposed based on a patient-specific approach by extracting undulated global and local features of preictal/ictal and interictal periods of EEG signals. The proposed method consists of feature extraction, classification, and regularization. The undulated global feature is extracted using phase correlation between two consecutive epochs of EEG signals and an undulated local feature is extracted using the fluctuation and deviation of EEG signals within the epoch. These features are further used for classification of preictal/ictal and interictal EEG signals. A regularization technique is applied on the classified outputs for the reduction of false alarms and improvement of the overall prediction accuracy (PA). The experimental results confirm that the proposed method provides high PA (i.e., 95.4%) with low false positive per hour using intracranial EEG signals in different brain locations of 21 patients from a benchmark dataset. Combining global and local features enables the transition point to be determined between different types of signals with greater accuracy, resulting successful versus unsuccessful prediction of seizure. The theoretical contribution of this study may provide an opportunity for the development of a clinical device to predict forthcoming seizure in real time.

Original languageEnglish
Article number7451217
Pages (from-to)208-217
Number of pages10
JournalIEEE Transactions on Biomedical Engineering
Volume64
Issue number1
Early online dateApr 2016
DOIs
Publication statusPublished - Jan 2017

Fingerprint

Electroencephalography
Feature extraction
Brain

Cite this

@article{3261009100804cbd8db72fe51fc9b518,
title = "Seizure Prediction using Undulated Global and Local Features",
abstract = "In this study, a seizure prediction method is proposed based on a patient-specific approach by extracting undulated global and local features of preictal/ictal and interictal periods of EEG signals. The proposed method consists of feature extraction, classification, and regularization. The undulated global feature is extracted using phase correlation between two consecutive epochs of EEG signals and an undulated local feature is extracted using the fluctuation and deviation of EEG signals within the epoch. These features are further used for classification of preictal/ictal and interictal EEG signals. A regularization technique is applied on the classified outputs for the reduction of false alarms and improvement of the overall prediction accuracy (PA). The experimental results confirm that the proposed method provides high PA (i.e., 95.4{\%}) with low false positive per hour using intracranial EEG signals in different brain locations of 21 patients from a benchmark dataset. Combining global and local features enables the transition point to be determined between different types of signals with greater accuracy, resulting successful versus unsuccessful prediction of seizure. The theoretical contribution of this study may provide an opportunity for the development of a clinical device to predict forthcoming seizure in real time.",
keywords = "Deviation, epilepsy, fluctuation, least square-support vector machine (LS-SVM), phase correlation, seizure",
author = "Parvez, {Mohammad Zavid} and Manoranjan Paul",
note = "Imported on 12 Apr 2017 - DigiTool details were: Journal title (773t) = IEEE Transactions on Biomedical Engineering. ISSNs: 0018-9294;",
year = "2017",
month = "1",
doi = "10.1109/TBME.2016.2553131",
language = "English",
volume = "64",
pages = "208--217",
journal = "IEEE Transactions on Bio-medical Electronics",
issn = "0018-9294",
publisher = "IEEE Computer Society",
number = "1",

}

Seizure Prediction using Undulated Global and Local Features. / Parvez, Mohammad Zavid; Paul, Manoranjan.

In: IEEE Transactions on Biomedical Engineering, Vol. 64, No. 1, 7451217, 01.2017, p. 208-217.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Seizure Prediction using Undulated Global and Local Features

AU - Parvez, Mohammad Zavid

AU - Paul, Manoranjan

N1 - Imported on 12 Apr 2017 - DigiTool details were: Journal title (773t) = IEEE Transactions on Biomedical Engineering. ISSNs: 0018-9294;

PY - 2017/1

Y1 - 2017/1

N2 - In this study, a seizure prediction method is proposed based on a patient-specific approach by extracting undulated global and local features of preictal/ictal and interictal periods of EEG signals. The proposed method consists of feature extraction, classification, and regularization. The undulated global feature is extracted using phase correlation between two consecutive epochs of EEG signals and an undulated local feature is extracted using the fluctuation and deviation of EEG signals within the epoch. These features are further used for classification of preictal/ictal and interictal EEG signals. A regularization technique is applied on the classified outputs for the reduction of false alarms and improvement of the overall prediction accuracy (PA). The experimental results confirm that the proposed method provides high PA (i.e., 95.4%) with low false positive per hour using intracranial EEG signals in different brain locations of 21 patients from a benchmark dataset. Combining global and local features enables the transition point to be determined between different types of signals with greater accuracy, resulting successful versus unsuccessful prediction of seizure. The theoretical contribution of this study may provide an opportunity for the development of a clinical device to predict forthcoming seizure in real time.

AB - In this study, a seizure prediction method is proposed based on a patient-specific approach by extracting undulated global and local features of preictal/ictal and interictal periods of EEG signals. The proposed method consists of feature extraction, classification, and regularization. The undulated global feature is extracted using phase correlation between two consecutive epochs of EEG signals and an undulated local feature is extracted using the fluctuation and deviation of EEG signals within the epoch. These features are further used for classification of preictal/ictal and interictal EEG signals. A regularization technique is applied on the classified outputs for the reduction of false alarms and improvement of the overall prediction accuracy (PA). The experimental results confirm that the proposed method provides high PA (i.e., 95.4%) with low false positive per hour using intracranial EEG signals in different brain locations of 21 patients from a benchmark dataset. Combining global and local features enables the transition point to be determined between different types of signals with greater accuracy, resulting successful versus unsuccessful prediction of seizure. The theoretical contribution of this study may provide an opportunity for the development of a clinical device to predict forthcoming seizure in real time.

KW - Deviation

KW - epilepsy

KW - fluctuation

KW - least square-support vector machine (LS-SVM)

KW - phase correlation

KW - seizure

UR - http://www.scopus.com/inward/record.url?scp=85008704699&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85008704699&partnerID=8YFLogxK

U2 - 10.1109/TBME.2016.2553131

DO - 10.1109/TBME.2016.2553131

M3 - Article

C2 - 27093309

VL - 64

SP - 208

EP - 217

JO - IEEE Transactions on Bio-medical Electronics

JF - IEEE Transactions on Bio-medical Electronics

SN - 0018-9294

IS - 1

M1 - 7451217

ER -