Applications of global optimization and dynamical systems to prediction of epileptic seizures

  • Authors:

  • W. Chaovalitwongse;P. M. Pardalos;L. D. Iasemidis;D.-S. Shiau;J. C. Sackellares

  • Affiliations:

  • Department of Industrial and Systems Engineering, University of Florida;Departments of Industrial and Systems Engineering, Computer Science, and Biomedical Engineering, University of Florida;Department of Biomedical Engineering, Arizona State University;Department of Neuroscience, University of Florida;Departments of Neuroscience, Neurology, and Biomedical Engineering, University of Florida, The Malcom Randall Department of Veterans Affairs Medical Center, Gainesville

  • Venue:
  • Quantitative neuroscience
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

Seizure occurrences seem to be random and unpredictable. However, recent studies in epileptic patients suggest that seizures are deterministic rather than random. There is growing evidence that seizures develop minutes to hours before clinical onset. Our previous studies have shown that quantitative analysis based on chaos theory of long-term intracranial electroencephalogram (EEG) recordings may enable us to observe the seizure's development in advance before clinical onset. The period of seizure's development is called a preictal transition period, which is characterized by gradual dynamical changes in EEG signals of critical electrode sites from asymptomatic interictal state to seizure. Techniques used to detect a preictal transition include statistical analysis of EEG signals, optimization techniques, and nonlinear dynamics. In this paper, we herein present optimization techniques, specifically multi-quadratic 0-1 programming, for the selection of the cortical sites that are involved with seizure's development during the preictal transition period. The results of this study can be used as a criterion to preselect the critical electrode sites that can be used to predict epileptic seizures.