Chaotic Neuron Dynamics, Synchronization, and Feature Binding

Abstract

Neuroscience studies how a large collection of coupled neurons combines external data with internal memories into coherent patterns of meaning. Such a process is called “feature binding”, insofar as the coherent patterns combine together features which are extracted separately by specialized cells, but which do not make sense as isolated items. A powerful conjecture, with experimental confirmation, is that feature binding implies the mutual synchronization of axonal spike trains in neurons which can be far away and yet contribute to a well defined perception by sharing the same time code.

Based on recent investigations of homoclinic chaotic systems, and how they mutually synchronize, a novel conjecture on the dynamics of the single neuron is formulated. Homoclinic chaos implies the recurrent return of the dynamical trajectory to a saddle focus, in whose neighbourhood the system susceptibility (response to an external perturbation) is very high and hence it is very easy to lock to an external stimulus. Thus homoclinic chaos appears as the easiest way to encode information by a train of equal spikes occurring at erratic times.

In conventional measurements we read the number indicated by a meter’s pointer and assign to the measured object a set position corresponding to that number. On the contrary, a time code requires a decision time T sufficiently longer than the minimal interspike separation t ₁, so that the total number of different set elements is related in some way to the size $\frac{T}{t_1}$. In neuroscience it has been shown that T ~ 200ms while t ₁ ~ 3 ms .

In a sensory layer of the brain neocortex an external stimulus spreads over a large assembly of neurons building up a collective state, thus synchronization of trains of different individual neurons is the basis of a coherent perception.

The percept space can be given a metric structure by introducing a distance measure. This distance is conjugate of the duration time in the sense that an uncertainty relation is associated with time limited perceptions. If the synchronized train is truncated at a time ΔT < T, then the corresponding identification of a percept P carries an uncertainty cloud ΔP. As two uncertainty clouds overlap, interference occurs; this is a quantum behavior. Thus the quantum formalism is not exclusively limited to microscopic phenomena, but here it is conjectured to be the appropriate description of truncated perceptions.

This quantum feature is not related to Planck’s action but to the details of the perceptual chain.