|
University of Sydney
School of Mathematics and Statistics
Professor Peter Robinson
School of Physics, University of Sydney
Neurophysical Modelling of Brain Dynamics
Wednesday, May 30th, 2-3pm, Carslaw 275.
A recently developed neurophysical model of the generation of brain
electrical activity is outlined and applied to electroencephalograms
(EEGs) and evoked response potentials to stimuli (ERPs) in normal
subjects, and in subjects suffering from disorders including epilepsy.
The model incorporates both single-neuron physiology and the
large-scale anatomy of corticocortical and corticothalamic pathways,
including synaptic strengths, dendritic propagation, nonlinear firing
responses, and axonal conduction. Under this model, small
perturbations around steady-state conditions account for the spatial
and temporal observed EEGs as functions of state of arousal, and can
be analyzed using wave equations. Similarly for evoked response
potentials, which arise as impulse response functions of the system.
It is found that feedback via the thalamus is critical in determining
the forms of the EEG and evoked potentials, the transition between
sleep and waking, and the stability of the brain against seizures. A
number of common disorders correspond to significant changes in EEGs,
which can be quantified in terms of underlying physiology using
inverse modeling techniques within the framework of our theory. In
the nonlinear regime, limit-cycle and chaotic behavior are seen. In
particular, when negative feedback via the thalamus falls below a
threshold relative to positive feedback, limit cycle oscillations
develop with the characteristic 3 Hz spike-and-wave form of petit mal
epilepsy.
|