Hi, I'm interested in programming an integrate and fire neuron which is somewhat similar to intfire4. However, I want to work with two types of inhibitory synapses (GABAA fast and slow) and also add noise.
So, in order to construct such a cell, I need to know how can I reset a voltage (of the soma compartment, say), I read a material about state variable discontinuities, but couldn't find anything about programming a discontinuity in v.
Thank you.
Programming new integrate and fire neurons: how to reset v?
-
ted
- Site Admin
- Posts: 6384
- Joined: Wed May 18, 2005 4:50 pm
- Location: Yale University School of Medicine
- Contact:
v is reserved for membrane potential, which is computed by numerical integration of the
discretized cable equation. Artificial spiking cells ("integrate and fire cells") have a
"membrane state variable" that has nothing to do with the cable equation, numerical
integration, or v.
The membrane state variable of an artificial spiking cell must be described by an equation
that has an analytical solution. The case that you're interested in would be described by
at least four first-order ODEs:
one for the membrane state variable
one for the "excitatory input" state variable
one for the "fast inhibitory" state variable
one (or maybe two) for the "slow inhibitory" state variable
Your task is to come up with an analytical solution for this set of equations, so that, given
the values of all state variables at the present moment, one can compute, from algebraic
formulas, the values of all state variables at any time in the future (assuming that no new
input events arrive). And you have to come up with a reasonably efficient way to
discover if, and when, the membrane state variable is going to cross "firing threshold" at
some point in the future.
If you have not already done so, it may be useful to read chapter 10 of The NEURON
Book, or
Hines, M.L. and Carnevale, N.T.
Discrete event simulation in the NEURON environment.
Neurocomputing 58-60:1117-1122, 2004
which is downloadable from http://www.neuron.yale.edu/neuron/bib/nrnpubs.html
discretized cable equation. Artificial spiking cells ("integrate and fire cells") have a
"membrane state variable" that has nothing to do with the cable equation, numerical
integration, or v.
The membrane state variable of an artificial spiking cell must be described by an equation
that has an analytical solution. The case that you're interested in would be described by
at least four first-order ODEs:
one for the membrane state variable
one for the "excitatory input" state variable
one for the "fast inhibitory" state variable
one (or maybe two) for the "slow inhibitory" state variable
Your task is to come up with an analytical solution for this set of equations, so that, given
the values of all state variables at the present moment, one can compute, from algebraic
formulas, the values of all state variables at any time in the future (assuming that no new
input events arrive). And you have to come up with a reasonably efficient way to
discover if, and when, the membrane state variable is going to cross "firing threshold" at
some point in the future.
If you have not already done so, it may be useful to read chapter 10 of The NEURON
Book, or
Hines, M.L. and Carnevale, N.T.
Discrete event simulation in the NEURON environment.
Neurocomputing 58-60:1117-1122, 2004
which is downloadable from http://www.neuron.yale.edu/neuron/bib/nrnpubs.html