electroencephlogram database?

Rich Cooper richcooper1 at mindspring.com
Thu Jan 24 15:18:12 EST 2002

> > Each action potential is an impulse of about 1 ms duration, but there
> > is some theorizing that perhaps the molecule that crosses the synaptic
> > gap may have a message that depends on its specific structure.  If
> > so, then ....
> Different types of neurons release different molecules. Any single
> neuron may use more than one (2-3) different molecules as a signal.
> It is clear that the massage is completely dependent on the structure of
> the signal molecule(s).
> But do you in your question actually mean the case where one neuron
> may release different quantitative combinations of its two signal
> molecules in different circumstances or from its own various terminals?
> Dag Stenberg

The latter.  There was an article in Science News (long ago) that
indicated much more diversity in the molecule that moves across
the gap than previously thought.  The article mentioned vision
stimuli that were lines, surfaces, and other geometric artifacts,
and said that the way neurons responded to the stimuli was
dependent on the field, not on the point source that was directed
to that specific neuron.  They showed a Gray-code like image,
with black and white squares making up an image with quadrature
shapes.  Different quadrature shapes seemed to fire the same
neural spikes, but they indicated that the actual message might
be encoded in the molecule, not in whether there is a spike
or not.  They theorized about how DNA/RNA can encode
information in molecules with the same constituents, but with
different encodings of these constituents, and thought that
might explain the sensitivity to different quadrature shapes.

In the EE world, there has been work on time-of-flight signaling
techniques where coherence of a signal at a receiver depends
on the multiple paths it takes to get there.  The idea is that, with
a constant spread of two peaks in time, the receiver can delay
one peak a fixed amount and compare it with the next.  If the
two peaks are consistent after the relative delay, the receiver
declares a hit, otherwise it declares a miss.  By correlating
the impulses with "dithered" transmitted pulses, the receiver
can pick low signals out of noise.  The GPS technology is
a somewhat similar method of getting very low power signals
out of random noise.

So it seems to me that the brain can, and probably does,
use these same techniques in a different way.  The path that
a signal takes from a pattern of retinal rods to a single neuron
depends on the brain's wiring, on impinging related signals
that inhibit or enable the impulses, and other quasirandom
effects.  It seems more likely than not that the brain would
have some equivalent mechanism to quasirandom correlation
of recognizable signal impulses.


More information about the Neur-sci mailing list