synchronization without oscillations

Leslie Kay lmk2 at garnet.berkeley.edu
Tue Aug 1 12:55:47 EST 1995


In article <3v82h2$8oc at portal.gmu.edu>,
HARRY R. ERWIN <herwin at osf1.gmu.edu> wrote:
>I call this flash synchronization, and it is seen in Freeman's models and
>data. Skinner has relevant data as well. Leslie Kay may have developed a
>way of efficiently recognizing it in vivo. Bernard Doyon's papers address
>the underlying dynamics, and I intend to look at it in my dissertation. 
>
>BTW, the oscillations are generated by the inherent dynamics of metabotropic 
>glutamate synapses (Hayashi) and seem to be an endogenous process that is 
>modulated by dynamic interactions in the dendritic tree. 

First let me point out that the technique for examining non-random phase
plane relationships between different brain regions was developed and
implemented by Larry Lancaster, not by me.  It is called NECTAR
(Nonparametric Exact Contingency Table Association Routine). [Lancaster, L.
and Kay, L., CNS*95 proceedings; Lancaster,L., Kay, L., and Freeman,W.J.,
1995 WCNN proceedings].

What we have shown using this technique is that at behaviorally
characteristic periods separate anatomical structures in the rat
rhinencephalon enter a nonrandom attractor state associated with odor
identification.  There are coherent, but non-periodic, events which can be
shown (by standard coherence and phase analyses) to occur at
physiologically plausible delays from one structure to the other in both
the forward direction (olfactory bulb into the olfactory cortex and limbic
areas) and in the centrifugal direction (originating in the entorhinal
cortex and then after a delay in the bulb). This is not synchronization,
since it doesn't occur at the same time.  The results do not even require
that the waveforms in the two structures be linearly related, although in
many cases they are.  It may indicate driving of one structure by the
other, but appearance of oscillations (quasi-periodic) is irrelevant.  All
of this work has been done using the local field potential recorded with
depth electrodes. It does not represent synchronization of units....yet.

The original question addressed the issue of synchronization of _units_
within less than one cycle of oscillation, so that the oscillation is not
what "causes" the synchronization.  This is seen quite nicely at the
population level in the olfactory bulb (OB) _EEG_. Recordings from spatial
arrays indicate that while the activity during a "burst" is only
quasi-periodic, the "synchronization" is fast and occurs often within one
period of the dominant frequency. This has been described as a "state
change" (similar to a phase transition in a physical system) in several
papers.  Here are two of them:

Freeman, W.J. (1994). Neural mechanisms underlying destabilization of
cortex by sensory input. Physica D. 75: 151-164

Kay,L., Shimoide,K., and Freeman,W.J. (1995) Comparison of EEG Time Series
from Rat Olfactory System with Model Composed of Nonlinear Coupled
Oscillators, International Journal of Bifurcation and Chaos, Vol 5, No 3,
in press.

It is important not to confuse units with the field potential or the EEG.
Oscillations in the latter two are a population property.  Oscillations in
single unit firing may or may not be indicative of the population dynamics.
Synchronization by means of oscillation requires too long a time.  The
brain only has a few periods (typically less than 5) before the state of
the system changes.

Leslie Kay
lmk2 at garnet.berkeley.edu




More information about the Neur-sci mailing list