-
A single-compartment model of neurons is used in
the simulations; this provides a model of the net behavior of each
neuron.
-
Neurons are modeled with a Hodgkin-Huxley
formulation that has been modified to a reduced form (Rinzel 1985, Av Ron 1994).
-
The membrane model includes spike-generating sodium
channels as well as active potassium and calcium channels,
calcium-dependent potassium channels, and transient potassium channels.
-
The synaptic current is modeled using a
double-exponential function which allows for modeling realistic rise
and fall times (Getting 1989).
-
Only short latency synaptic interactions are
included in the simulations.
-
The simulations here were limited to small networks
of neurons (up to 36) containing both excitatory and inhibitory neurons
and connections.
-
Each neuron has a defined number of excitatory and
inhibitory synaptic inputs from randomly chosen presynaptic inhibitory
and excitatory neurons (Figure 1).
-
A few neurons (< 30% of neurons in a network)
received
-
random, uncorrelated excitatory noise during the
simulations.