Next: Results Up:
index Previous: Introduction
Our model incorporates: 1) a spatially distributed network (NETWORK Fig.1)
capable of reproducing the spread of bursting activity and 2) a small external
network (FOCUS), which triggers the activity in the NETWORK. The FOCUS is
comprised of neurons capable of generating endogenous bursts (focus I) or
has the built-in structures of connections capable of maintaining self-sustained
oscillations (focus II). We also simulate lesion in a network (focus III)
by introducing to the NETWORK a sub-population of neurons with modified cell
membrane or network connections properties.
NETWORK is simulated as an array of 24 by 24 sub-networks, where each sub-network
is composed of 81 excitatory and 9 inhibitory neurons. This network architecture
resembles columnar organization in cortex. Neurons are modeled using a conductance-based
model. The excitatory neuron fire repetitive bursts in response to constant
current or synaptic input with a frequency dependent on the calcium elimination
rate. The inhibitory interneurons respond to excitatory input with high-frequency
nonadapting spike trains. Within a column each neuron has two excitatory
and two inhibitory synapses in input. The synaptic delays have a mean value
of 1.6 ms and standard deviation of 0.5 ms. Randomly selected excitatory
neurons in neighboring columns are connected. There are only excitatory connections
between columns and the total number of connections between two columns does
not exceed 16.
Next: Results Up:
index Previous: Introduction