Detection of seizures from small samples using nonlinear dynamic system theory

Ilker Yaylali, Huseyin Kocak, Prasanna Jayakar

Research output: Contribution to journalArticle

56 Citations (Scopus)

Abstract

The electroencephalogram (EEG), like many other biological phenomena, is quite likely governed by nonlinear dynamics. Certain characteristics of the underlying dynamics have recently been quantified by computing the correlation dimensions (D2) of EEG time series data. In this paper, D2 of the unbiased autocovariance function of the scalp EEG data was used to detect electrographic seizure activity. Digital EEG data were acquired at a sampling rate of 200 Hz per channel and organized in continuous frames (duration 2.56 s, 512 data points). To increase the reliability of D2 computations with short duration data, raw EEG data were initially simplified using unbiased autocovariance analysis to highlight the periodic activity that is present during seizures. The D2 computation was then performed from the unbiased autocovariance function of each channel using the Grassberger-Procaccia method with Theiler's box-assisted correlation algorithm. Even with short duration data, this preprocessing proved to be computationally robust and displayed no significant sensitivity to implementation details such as the choices of embedding dimension and box size. The system successfully identified various types of seizures in clinical studies.

Original languageEnglish (US)
Pages (from-to)743-751
Number of pages9
JournalIEEE Transactions on Biomedical Engineering
Volume43
Issue number7
DOIs
StatePublished - Jul 1996

Fingerprint

System theory
Electroencephalography
Dynamical systems
Time series
Sampling

ASJC Scopus subject areas

  • Biomedical Engineering

Cite this

Detection of seizures from small samples using nonlinear dynamic system theory. / Yaylali, Ilker; Kocak, Huseyin; Jayakar, Prasanna.

In: IEEE Transactions on Biomedical Engineering, Vol. 43, No. 7, 07.1996, p. 743-751.

Research output: Contribution to journalArticle

@article{3beea12d373d4fed854694bb3322c2b1,
title = "Detection of seizures from small samples using nonlinear dynamic system theory",
abstract = "The electroencephalogram (EEG), like many other biological phenomena, is quite likely governed by nonlinear dynamics. Certain characteristics of the underlying dynamics have recently been quantified by computing the correlation dimensions (D2) of EEG time series data. In this paper, D2 of the unbiased autocovariance function of the scalp EEG data was used to detect electrographic seizure activity. Digital EEG data were acquired at a sampling rate of 200 Hz per channel and organized in continuous frames (duration 2.56 s, 512 data points). To increase the reliability of D2 computations with short duration data, raw EEG data were initially simplified using unbiased autocovariance analysis to highlight the periodic activity that is present during seizures. The D2 computation was then performed from the unbiased autocovariance function of each channel using the Grassberger-Procaccia method with Theiler's box-assisted correlation algorithm. Even with short duration data, this preprocessing proved to be computationally robust and displayed no significant sensitivity to implementation details such as the choices of embedding dimension and box size. The system successfully identified various types of seizures in clinical studies.",
author = "Ilker Yaylali and Huseyin Kocak and Prasanna Jayakar",
year = "1996",
month = "7",
doi = "10.1109/10.503182",
language = "English (US)",
volume = "43",
pages = "743--751",
journal = "IEEE Transactions on Biomedical Engineering",
issn = "0018-9294",
publisher = "IEEE Computer Society",
number = "7",

}

TY - JOUR

T1 - Detection of seizures from small samples using nonlinear dynamic system theory

AU - Yaylali, Ilker

AU - Kocak, Huseyin

AU - Jayakar, Prasanna

PY - 1996/7

Y1 - 1996/7

N2 - The electroencephalogram (EEG), like many other biological phenomena, is quite likely governed by nonlinear dynamics. Certain characteristics of the underlying dynamics have recently been quantified by computing the correlation dimensions (D2) of EEG time series data. In this paper, D2 of the unbiased autocovariance function of the scalp EEG data was used to detect electrographic seizure activity. Digital EEG data were acquired at a sampling rate of 200 Hz per channel and organized in continuous frames (duration 2.56 s, 512 data points). To increase the reliability of D2 computations with short duration data, raw EEG data were initially simplified using unbiased autocovariance analysis to highlight the periodic activity that is present during seizures. The D2 computation was then performed from the unbiased autocovariance function of each channel using the Grassberger-Procaccia method with Theiler's box-assisted correlation algorithm. Even with short duration data, this preprocessing proved to be computationally robust and displayed no significant sensitivity to implementation details such as the choices of embedding dimension and box size. The system successfully identified various types of seizures in clinical studies.

AB - The electroencephalogram (EEG), like many other biological phenomena, is quite likely governed by nonlinear dynamics. Certain characteristics of the underlying dynamics have recently been quantified by computing the correlation dimensions (D2) of EEG time series data. In this paper, D2 of the unbiased autocovariance function of the scalp EEG data was used to detect electrographic seizure activity. Digital EEG data were acquired at a sampling rate of 200 Hz per channel and organized in continuous frames (duration 2.56 s, 512 data points). To increase the reliability of D2 computations with short duration data, raw EEG data were initially simplified using unbiased autocovariance analysis to highlight the periodic activity that is present during seizures. The D2 computation was then performed from the unbiased autocovariance function of each channel using the Grassberger-Procaccia method with Theiler's box-assisted correlation algorithm. Even with short duration data, this preprocessing proved to be computationally robust and displayed no significant sensitivity to implementation details such as the choices of embedding dimension and box size. The system successfully identified various types of seizures in clinical studies.

UR - http://www.scopus.com/inward/record.url?scp=0030200521&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030200521&partnerID=8YFLogxK

U2 - 10.1109/10.503182

DO - 10.1109/10.503182

M3 - Article

C2 - 9216146

AN - SCOPUS:0030200521

VL - 43

SP - 743

EP - 751

JO - IEEE Transactions on Biomedical Engineering

JF - IEEE Transactions on Biomedical Engineering

SN - 0018-9294

IS - 7

ER -