Home
Contact
Short CV
Research
Publications
Software
Links
Impressum
University of Plymouth | Centre for Theoret. and Comput. Neuroscience | Home
A network of integrate and fire units
"Integrate and fire units" are very simple models of neurons in the nervous system. They are described by a single variable, their "potential", and consist of a low-pass, which models passive membrane properties of a neuron, and a spike encoder, which emits a pulse every time the potential crosses a threshold of (say) 1. Afterwards the potential is discontinuously reset to 0 (the so-called "resting level"). In a network pulses of a neuron are distributed to other neurons, where they evoke an increase (excitation) or decrease (inhibition) in the potential of a target cell according to the sign and weight of the respective connection (also called "synapse").The simulation shows a 2 dimensional grid of integrate and fire neurons, where the potential values are color-coded and neighbouring neurons interact with each other via synapses with a certain strength. Each neuron further receives a certain input noise and a fixed (constant) input, which determines the fundamental firing frequency of the cells. If a neuron fires it is displayed in red for one time step.
According to the parameter settings, the firing of pulses of the neurons can be independent (vanishing synaptic strengths) or more or less synchronous (low noise, high synaptic strengths). In larger networks (try changing N and the size parameter) you can observe wave like patterns, where neurons are only locally synchronized. If the coupling strength is initially 0 you should set it to a small value, e.g., coupling=0.1
© 2004 by -thowe-