Predicting epileptic seizures in scalp EEG based on a variational bayesian gaussian mixture model of zero-crossing intervals

Ali Shahidi Zandi, Reza Tafreshi, Manouchehr Javidan, Guy A. Dumont

Research output: Contribution to journalArticle

50 Citations (Scopus)

Abstract

A novel patient-specific seizure prediction method based on the analysis of positive zero-crossing intervals in scalp electroencephalogram (EEG) is proposed. In a moving-window analysis, the histogram of these intervals for the current EEG epoch is computed, and the values corresponding to specific bins are selected as an observation. Then, the set of observations from the last 5 min is compared with two reference sets of data points (preictal and interictal) through novel measures of similarity and dissimilarity based on a variational Bayesian Gaussian mixture model of the data. A combined index is then computed and compared with a patient-specific threshold, resulting in a cumulative measure which is utilized to form an alarm sequence for each channel. Finally, this channel-based information is used to generate a seizure prediction alarm. The proposed method was evaluated using ∼561 h of scalp EEG including a total of 86 seizures in 20 patients. A high sensitivity of 88.34\% was achieved with a false prediction rate of 0.155 h-1 and an average prediction time of 22.5 min for the test dataset. The proposed method was also tested against a Poisson-based random predictor.

Original languageEnglish
Article number6400235
Pages (from-to)1401-1413
Number of pages13
JournalIEEE Transactions on Biomedical Engineering
Volume60
Issue number5
DOIs
Publication statusPublished - 2013

Fingerprint

Electroencephalography
Bins

Keywords

  • Electroencephalogram (EEG)
  • epileptic seizure prediction
  • variational Gaussian mixture model (GMM)
  • zero-crossing intervals

ASJC Scopus subject areas

  • Biomedical Engineering

Cite this

Predicting epileptic seizures in scalp EEG based on a variational bayesian gaussian mixture model of zero-crossing intervals. / Zandi, Ali Shahidi; Tafreshi, Reza; Javidan, Manouchehr; Dumont, Guy A.

In: IEEE Transactions on Biomedical Engineering, Vol. 60, No. 5, 6400235, 2013, p. 1401-1413.

Research output: Contribution to journalArticle

@article{2bea41b91238466d97a0414c210c0dc8,
title = "Predicting epileptic seizures in scalp EEG based on a variational bayesian gaussian mixture model of zero-crossing intervals",
abstract = "A novel patient-specific seizure prediction method based on the analysis of positive zero-crossing intervals in scalp electroencephalogram (EEG) is proposed. In a moving-window analysis, the histogram of these intervals for the current EEG epoch is computed, and the values corresponding to specific bins are selected as an observation. Then, the set of observations from the last 5 min is compared with two reference sets of data points (preictal and interictal) through novel measures of similarity and dissimilarity based on a variational Bayesian Gaussian mixture model of the data. A combined index is then computed and compared with a patient-specific threshold, resulting in a cumulative measure which is utilized to form an alarm sequence for each channel. Finally, this channel-based information is used to generate a seizure prediction alarm. The proposed method was evaluated using ∼561 h of scalp EEG including a total of 86 seizures in 20 patients. A high sensitivity of 88.34\{\%} was achieved with a false prediction rate of 0.155 h-1 and an average prediction time of 22.5 min for the test dataset. The proposed method was also tested against a Poisson-based random predictor.",
keywords = "Electroencephalogram (EEG), epileptic seizure prediction, variational Gaussian mixture model (GMM), zero-crossing intervals",
author = "Zandi, {Ali Shahidi} and Reza Tafreshi and Manouchehr Javidan and Dumont, {Guy A.}",
year = "2013",
doi = "10.1109/TBME.2012.2237399",
language = "English",
volume = "60",
pages = "1401--1413",
journal = "IEEE Transactions on Biomedical Engineering",
issn = "0018-9294",
publisher = "IEEE Computer Society",
number = "5",

}

TY - JOUR

T1 - Predicting epileptic seizures in scalp EEG based on a variational bayesian gaussian mixture model of zero-crossing intervals

AU - Zandi, Ali Shahidi

AU - Tafreshi, Reza

AU - Javidan, Manouchehr

AU - Dumont, Guy A.

PY - 2013

Y1 - 2013

N2 - A novel patient-specific seizure prediction method based on the analysis of positive zero-crossing intervals in scalp electroencephalogram (EEG) is proposed. In a moving-window analysis, the histogram of these intervals for the current EEG epoch is computed, and the values corresponding to specific bins are selected as an observation. Then, the set of observations from the last 5 min is compared with two reference sets of data points (preictal and interictal) through novel measures of similarity and dissimilarity based on a variational Bayesian Gaussian mixture model of the data. A combined index is then computed and compared with a patient-specific threshold, resulting in a cumulative measure which is utilized to form an alarm sequence for each channel. Finally, this channel-based information is used to generate a seizure prediction alarm. The proposed method was evaluated using ∼561 h of scalp EEG including a total of 86 seizures in 20 patients. A high sensitivity of 88.34\% was achieved with a false prediction rate of 0.155 h-1 and an average prediction time of 22.5 min for the test dataset. The proposed method was also tested against a Poisson-based random predictor.

AB - A novel patient-specific seizure prediction method based on the analysis of positive zero-crossing intervals in scalp electroencephalogram (EEG) is proposed. In a moving-window analysis, the histogram of these intervals for the current EEG epoch is computed, and the values corresponding to specific bins are selected as an observation. Then, the set of observations from the last 5 min is compared with two reference sets of data points (preictal and interictal) through novel measures of similarity and dissimilarity based on a variational Bayesian Gaussian mixture model of the data. A combined index is then computed and compared with a patient-specific threshold, resulting in a cumulative measure which is utilized to form an alarm sequence for each channel. Finally, this channel-based information is used to generate a seizure prediction alarm. The proposed method was evaluated using ∼561 h of scalp EEG including a total of 86 seizures in 20 patients. A high sensitivity of 88.34\% was achieved with a false prediction rate of 0.155 h-1 and an average prediction time of 22.5 min for the test dataset. The proposed method was also tested against a Poisson-based random predictor.

KW - Electroencephalogram (EEG)

KW - epileptic seizure prediction

KW - variational Gaussian mixture model (GMM)

KW - zero-crossing intervals

UR - http://www.scopus.com/inward/record.url?scp=84876752338&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84876752338&partnerID=8YFLogxK

U2 - 10.1109/TBME.2012.2237399

DO - 10.1109/TBME.2012.2237399

M3 - Article

VL - 60

SP - 1401

EP - 1413

JO - IEEE Transactions on Biomedical Engineering

JF - IEEE Transactions on Biomedical Engineering

SN - 0018-9294

IS - 5

M1 - 6400235

ER -