APPLICATION OF NON-LINEAR DYNAMICS FOR PREDICTING EPILEPTIC SEIZURES
Aspects of non-linear dynamics were applied with moderate success in an attempt to predict epileptic seizures based on the maximum Lyapunov exponents of the patients' electroencephalograms (EEG's). The basis of all of the techniques used is two point correlations between the Lyapunov exponents of the time-varying voltages of the electrodes used to record the EEG's. Among the calculations made were the relative distances of the electrodes from the focus of the seizures and the integrals of the correlations. These techniques enable impending seizures to be seen qualitatively an hour or two in advance in some cases, but still need to be quantified in a reliable algorithm that will predict weak seizures without giving too many false alarms. Attempts were also made to draws parallels with the two dimensional Ising model of magnetism by treating the occurrence of a seizure as a phase transition.
Keywords: non-linear dynamics, epileptic seizures, Lyapunov exponents
Article MetricsAbstract view : 565 times
PDF - 203 times
- There are currently no refbacks.