Joining the benefits

Combining epileptic seizure prediction methods

Hinnerk Feldwisch-Drentrup, Bjoern Schelter, Michael Jachan, Jakob Nawrath, Jens Timmer, Andreas Schulze-Bonhage

Research output: Contribution to journalArticle

55 Citations (Scopus)

Abstract

P>Purpose:

In recent years, a variety of methods developed in the field of linear and nonlinear time series analysis have been used to obtain reliable predictions of epileptic seizures. Because individual methods for seizure prediction so far have shown statistical significance but insufficient performance for clinical applications, we investigated possible improvements by combining algorithms capturing different aspects of electroencephalogram (EEG) dynamics.

Methods:

We applied the mean phase coherence and the dynamic similarity index to long-term continuous intracranial EEG data. The predictive performance of both methods was assessed and statistically evaluated separately, as well as by using logical "AND" and "OR" combinations.

Results:

Used independently, either method resulted in a statistically significant prediction performance in only a few patients. Particularly the "AND" combination led to improved prediction performances, leading to an increase in sensitivity and/or specificity. For a maximum false prediction rate of 0.15/h, the mean sensitivity improved from about 25% for the individual methods to 43.2% for the "AND" and to 35.2% for the "OR" combination.

Discussion:

This study shows that combinations of prediction methods are promising new approaches to enhance seizure prediction performance considerably. It allows merging the individual benefits of prediction methods in a complementary manner. Because either sensitivity or specificity of seizure prediction methods can be improved depending on the needs of the desired clinical application, the combination opens a new window for future use in a clinical setting.

Original languageEnglish
Pages (from-to)1598-1606
Number of pages9
JournalEpilepsia
Volume51
Issue number8
Early online date7 Jan 2010
DOIs
Publication statusPublished - Aug 2010

Keywords

  • epilepsy
  • seizure prediction
  • prediction method combination

Cite this

Feldwisch-Drentrup, H., Schelter, B., Jachan, M., Nawrath, J., Timmer, J., & Schulze-Bonhage, A. (2010). Joining the benefits: Combining epileptic seizure prediction methods. Epilepsia, 51(8), 1598-1606. https://doi.org/10.1111/j.1528-1167.2009.02497.x

Joining the benefits : Combining epileptic seizure prediction methods. / Feldwisch-Drentrup, Hinnerk; Schelter, Bjoern; Jachan, Michael; Nawrath, Jakob; Timmer, Jens; Schulze-Bonhage, Andreas.

In: Epilepsia, Vol. 51, No. 8, 08.2010, p. 1598-1606.

Research output: Contribution to journalArticle

Feldwisch-Drentrup, H, Schelter, B, Jachan, M, Nawrath, J, Timmer, J & Schulze-Bonhage, A 2010, 'Joining the benefits: Combining epileptic seizure prediction methods', Epilepsia, vol. 51, no. 8, pp. 1598-1606. https://doi.org/10.1111/j.1528-1167.2009.02497.x
Feldwisch-Drentrup H, Schelter B, Jachan M, Nawrath J, Timmer J, Schulze-Bonhage A. Joining the benefits: Combining epileptic seizure prediction methods. Epilepsia. 2010 Aug;51(8):1598-1606. https://doi.org/10.1111/j.1528-1167.2009.02497.x
Feldwisch-Drentrup, Hinnerk ; Schelter, Bjoern ; Jachan, Michael ; Nawrath, Jakob ; Timmer, Jens ; Schulze-Bonhage, Andreas. / Joining the benefits : Combining epileptic seizure prediction methods. In: Epilepsia. 2010 ; Vol. 51, No. 8. pp. 1598-1606.
@article{a2d2c2644dde4c5e94c97b103169f7ce,
title = "Joining the benefits: Combining epileptic seizure prediction methods",
abstract = "P>Purpose:In recent years, a variety of methods developed in the field of linear and nonlinear time series analysis have been used to obtain reliable predictions of epileptic seizures. Because individual methods for seizure prediction so far have shown statistical significance but insufficient performance for clinical applications, we investigated possible improvements by combining algorithms capturing different aspects of electroencephalogram (EEG) dynamics.Methods:We applied the mean phase coherence and the dynamic similarity index to long-term continuous intracranial EEG data. The predictive performance of both methods was assessed and statistically evaluated separately, as well as by using logical {"}AND{"} and {"}OR{"} combinations.Results:Used independently, either method resulted in a statistically significant prediction performance in only a few patients. Particularly the {"}AND{"} combination led to improved prediction performances, leading to an increase in sensitivity and/or specificity. For a maximum false prediction rate of 0.15/h, the mean sensitivity improved from about 25{\%} for the individual methods to 43.2{\%} for the {"}AND{"} and to 35.2{\%} for the {"}OR{"} combination.Discussion:This study shows that combinations of prediction methods are promising new approaches to enhance seizure prediction performance considerably. It allows merging the individual benefits of prediction methods in a complementary manner. Because either sensitivity or specificity of seizure prediction methods can be improved depending on the needs of the desired clinical application, the combination opens a new window for future use in a clinical setting.",
keywords = "epilepsy, seizure prediction, prediction method combination",
author = "Hinnerk Feldwisch-Drentrup and Bjoern Schelter and Michael Jachan and Jakob Nawrath and Jens Timmer and Andreas Schulze-Bonhage",
year = "2010",
month = "8",
doi = "10.1111/j.1528-1167.2009.02497.x",
language = "English",
volume = "51",
pages = "1598--1606",
journal = "Epilepsia",
issn = "0013-9580",
publisher = "Wiley-Blackwell",
number = "8",

}

TY - JOUR

T1 - Joining the benefits

T2 - Combining epileptic seizure prediction methods

AU - Feldwisch-Drentrup, Hinnerk

AU - Schelter, Bjoern

AU - Jachan, Michael

AU - Nawrath, Jakob

AU - Timmer, Jens

AU - Schulze-Bonhage, Andreas

PY - 2010/8

Y1 - 2010/8

N2 - P>Purpose:In recent years, a variety of methods developed in the field of linear and nonlinear time series analysis have been used to obtain reliable predictions of epileptic seizures. Because individual methods for seizure prediction so far have shown statistical significance but insufficient performance for clinical applications, we investigated possible improvements by combining algorithms capturing different aspects of electroencephalogram (EEG) dynamics.Methods:We applied the mean phase coherence and the dynamic similarity index to long-term continuous intracranial EEG data. The predictive performance of both methods was assessed and statistically evaluated separately, as well as by using logical "AND" and "OR" combinations.Results:Used independently, either method resulted in a statistically significant prediction performance in only a few patients. Particularly the "AND" combination led to improved prediction performances, leading to an increase in sensitivity and/or specificity. For a maximum false prediction rate of 0.15/h, the mean sensitivity improved from about 25% for the individual methods to 43.2% for the "AND" and to 35.2% for the "OR" combination.Discussion:This study shows that combinations of prediction methods are promising new approaches to enhance seizure prediction performance considerably. It allows merging the individual benefits of prediction methods in a complementary manner. Because either sensitivity or specificity of seizure prediction methods can be improved depending on the needs of the desired clinical application, the combination opens a new window for future use in a clinical setting.

AB - P>Purpose:In recent years, a variety of methods developed in the field of linear and nonlinear time series analysis have been used to obtain reliable predictions of epileptic seizures. Because individual methods for seizure prediction so far have shown statistical significance but insufficient performance for clinical applications, we investigated possible improvements by combining algorithms capturing different aspects of electroencephalogram (EEG) dynamics.Methods:We applied the mean phase coherence and the dynamic similarity index to long-term continuous intracranial EEG data. The predictive performance of both methods was assessed and statistically evaluated separately, as well as by using logical "AND" and "OR" combinations.Results:Used independently, either method resulted in a statistically significant prediction performance in only a few patients. Particularly the "AND" combination led to improved prediction performances, leading to an increase in sensitivity and/or specificity. For a maximum false prediction rate of 0.15/h, the mean sensitivity improved from about 25% for the individual methods to 43.2% for the "AND" and to 35.2% for the "OR" combination.Discussion:This study shows that combinations of prediction methods are promising new approaches to enhance seizure prediction performance considerably. It allows merging the individual benefits of prediction methods in a complementary manner. Because either sensitivity or specificity of seizure prediction methods can be improved depending on the needs of the desired clinical application, the combination opens a new window for future use in a clinical setting.

KW - epilepsy

KW - seizure prediction

KW - prediction method combination

U2 - 10.1111/j.1528-1167.2009.02497.x

DO - 10.1111/j.1528-1167.2009.02497.x

M3 - Article

VL - 51

SP - 1598

EP - 1606

JO - Epilepsia

JF - Epilepsia

SN - 0013-9580

IS - 8

ER -