I wonder if anyone can help me with something I need to know for some of
the simulations I am currently trying to configure.
If I assume that a neuron has of the order of a thousand inputs and that
it is possible for a single input across a synapse with minimal efficacy
to cause the neuron to fire.
Which seems reasonable because it would seem necessary for neurons to be
able to eventually detect and fire upon continual activity of a single
but weak input.
Then - now for the questions -
How many reasonable inputs (2/3 maximum firing rate) across synapses
with reasonable efficacy (2/3 maximum efficacy) are needed to
produce a reasonable output from the target neuron?
How many reasonable inputs across synapses with reasonable efficacy
are need to saturate a neuron?
[ Of course, I understand there is great variation within the
nervous system. Exact answers aren't necessary, just some
idea of the scale of things would be most helpful. ]
The reason I ask this is that there would seem to be considerable varia-
tion between the postsynaptic membrane depolarization caused by 1 neuron
and that caused by a 1000 neurons firing. How do neurons cope and discr-
iminate between these cases.
A solution to the above is to make the firing of the neuron a stochastic
function of the input. In this way, a small depolarization from a single
input most likely won't make the neuron fire, but once in a while will (
giving it the chance to eventually increase its synaptic efficacy).
I would be most grateful if anyone having any comments or references to
papers could please e-mail me as well as posting, as I don't get to read
this group too regularly.
Thanx in Advance,
University of NSW.
E-MAIL : ashley at cs.unsw.oz.au AARNet
Academic Address: Residential Address:
Schools of EE and CS&E, (AI Lab) c/o Basser College, (Flat 7A)
University of New South Wales, The Kensington Colleges,
Box 1,PO KENSINGTON,N.S.W.,2033. Box 24,PO KENSINGTON,N.S.W,2033.
Ph. (02) 697-3940 Fx. (02) 663-4576 Ph. (02) 663-8117 (Preferred!)