×

zbMATH — the first resource for mathematics

Epilepsy as a dynamic disease. (English) Zbl 1058.92027
Biological and Medical Physics Series. Berlin: Springer (ISBN 3-540-42762-7/hbk). xxxii, 417 p. (2003).
The volume is an unprecedented collection of articles whose overall theme is to discuss the various aspects involved in the design of an implantable brain defibrilator as a therapy for epilepsy. It is written by an international multidisciplinary team of experts ranging from biologists and surgeons to mathematicians and engineers. The text is a full treatment of the subject starting from the clinical aspects of the disease, delving into the issues of neuronal synchronization and propagation of seizures and exploring the subjects of seizure prediction and feedback control of epileptic seizures.
The first four chapters examine the clinical and surgical aspects of seisures arising in the temporal lobe complex, the type of epilepsy that has been most extensively studied. The underlying thesis is that epilepsy is a dynamically changing disease which develops as seizures arising in the temporal lobe gradually involving the whole cortex. In the words of the authors, “an epileptic system is a spatially distributed system of groups of neurons located in the deep regions of the brain which cooperate to control the onset, propagation and arrest of epileptic seizures”. To develop a brain defibrilator the routes of seisure propagation must be well understood. Three possible pathways of seizure generalization are assessed by comparing their propagation velocities. It is concluded that the most probable pathway is the spread via reciprocal connections between neurons in the gray matter and the subcortical and brainstem nuclei. A review of the results of depth electrode recordings suggests the site of the effector arm of a brain stimulator. Evidence reviewed in the fourth chapter leads to the conclusion that a specific cell type, the astrocyte, possibly plays a central role in the maintenance and development of the epileptic focus.
As a seizure represents synchronization of a massively large population of neurons, the next four chapters are dedicated to the topic of neural synchronization. The electro-encephalogram (EEG) is the main tool for monitoring epilepsy. The authors postulate that the generators of the EEG are the postsynaptic potentials that summate on the neural surface and the EEG is a measure of neural synchrony. Three basic patterns of neural synchrony associated with temporal lobe epilepsy observed by examining EEG are discussed. It is suggested that strategies that alter neural synchrony are the most effective way to stop a seizure. The value of coherence analysis of EEG data is demonstrated on data from two patients and it is hypothesised that EEG coherence may provide a method to target the epileptic network. Mathematical models of the synchronization of populations of spiking neurons as a function of synaptic coupling strength and the theory of weak coupling synchronization are reviewed in chapter 7. Chapter 8 is dedicated to studies of the effects of periodic stimulation on populations of neurons from the standpoint of nonlinear dynamics. The observations in this chapter support the possibility to construct a brain defibrilator based on periodic stimulation.
More mathematical observations from the theory of excitable media are included in the first of the next three chapters dedicated to the propagation of seizures. The chapter discusses the formation and properties of travelling spiral waves, and the simplest form of self-maintained wave and propagation. Mechanisms to disrupt spiral waves are proposed as a message to the future developer that strategies to alter the propagation of neural activity waves may be important for the design of the defibrilator. As the theory of excitable media has been applied in cardiac defibrilation, the next chapter 10 discusses the possible analogies between wave propagation in cardiac and neural tissues. The final chapter in this series draws analogies to pattern formation in microbial populations to alert investigators to the importance of identifying the mechanism which lies behind the appearance of new spatiotemporal dynamics in biological excitable media.
Chapters 12 and 13 are dedicated to the subject of seizure prediction and the possibility to perform the prediction 2-3 minutes before the seizure. The texts review the progress in seizure prediction from time series analysis of the EEG. The first chapter discusses the development of time series methods based on the properties of nonlinear dynamical systems. It is stated that time series methods based on cross-correlation approache to detect non-autonomy of the EEG are more robust for seizure prediction than ones based on the correlation integral. The second chapter examines the ability of computerized time series analysis to predict seizure occurrence relative to that of a trained EEG reader. Five methods of linear time series analysis and two methods of nonlinear time series analysis are compared with regard to reliability and speed of detection. Some advantages of the linear methods over the nonlinear ones are supported and it is concluded that a lead time of 2-3 minutes before seizure might be sufficient for the activation of a brain defibrilator.
The last part of the book is dedicated to the feedback control of the epileptic seizures. The rationale and outcomes of the earliest clinical attempts to control seizures using deep brain stimulation performed by a group led by Professor S.A. Chkhenkeli in Tbilisi, Georgian Republic, are discussed in chapter 14. The studies implicate that the stimulation site is likely more important than the details of the stimulus protocol. The next chapter describes the development of the first device capable of performing feedback control of seizures. Preliminary results from applications of the device to rat hippocampal slices suggest that electric fields smaller than those emitted by a cellphone may be effective. Because of the evidence that a brief electrical stimulus can sometimes stop a seizure if given before the onset of the event, chapter 16 discusses the conjecture that epilepsy arises in a multistable dynamical system. Specifically the expected dynamics of an EEG as a complex multistable noisy dynamical system with retarded variables is considered. Such systems exhibit statistical periodicity: a phenomenon whose probability density cycles periodically. The message of this chapter is that the precise time of application of a stimulus by a defibrilator is an important factor. Chapter 17 discusses statistical methods for chaos control techniques and their successful application to epileptic seizures in rats. The final chapter reviews the technical challenges in the development of a brain defibrilator. It is concluded that the two most important problems in the construction of a brain defibrilator are a) its computational requirements and b) its biocompatibility.
This unique book lays the theoretical foundations of a much needed device for the treatment of epilepsy. It is of interest to a broad scientific community including medical scientists, engineers, biophysicists and mathematicians.

MSC:
92C50 Medical applications (general)
92C20 Neural biology
92-02 Research exposition (monographs, survey articles) pertaining to biology
92C55 Biomedical imaging and signal processing
37N25 Dynamical systems in biology
93D15 Stabilization of systems by feedback
92-06 Proceedings, conferences, collections, etc. pertaining to biology
00B15 Collections of articles of miscellaneous specific interest
PDF BibTeX XML Cite