question: job of a neuron - wavelets revisited...
meshinder at aol.com
Thu Sep 13 15:17:46 EST 2001
I'm still trying to understand this better. The temporal code of spikes carried
on a given neuron possiblly carries information, as does the spatiotemporal
pattern of spike times between neurons (see work of Gerstein, Aertsen, Fetz,
Deadwyler, McNaughton and the like). Also information is coded in the temporal
wave of activity as can be demonstrated by peristimulus histograms and
prediction functions on neuron firing in most regions of the brain where such
efforts have been applied.
So waves may be composed of patterns, and as waves travel around in highly
structured neuronal pathways from region to region they create more waves and
patterns. To interpret all this their seems like there would have to be an
organizing principle. The brain is not homogenous or random, but not rigid or
absolutely genetic in all of its architecture either.
What is the current thinking regarding how regionalized emmergent functions
develop into larger or meta-functions? I can conceptualize how patterns can
regulate processing through gating and state functions, but how would waves do
this? I understand how waves could enhance binding and input decomposition, but
could a pattern not constituting a wave do the same?
I don't mean to get into the particle/wave conundrum, and I may be missing some
aspect or perspective on this, so any help would be appreciated.
More information about the Neur-sci