Time as coding space?

https://doi.org/10.1016/S0959-4388(99)80026-9Get rights and content

Abstract

There is increasing evidence that neuronal networks can operate with a temporal resolution in the millisecond range. In principle, this provides the option to encode information not only in the amplitude of neuronal responses but also in the precise temporal relations between the discharges of distributed neurons.

Introduction

In investigations of neuronal correlates of behaviour, the measured neuronal variable is typically the stimulus- or action-related modulation of the discharge rate of neurons. This approach has led to an impressive body of evidence relating variations in the discharge rate of individual neurons to specific cognitive and motor processes. However, unless these variations of discharge rate are tightly time-locked to stimuli or motor events, this approach precludes assessment of information that could, in principle, be encoded in the precise temporal relations between the discharges of ensembles of neurons that contribute collectively to a particular function. Such information can be extracted only from multi-cell recordings and correlation analysis of simultaneously registered neuronal responses.

A steadily increasing number of studies has focused on the temporal properties of neuronal responses (reviewed in 1, 2, 3). These follow the discovery of precise synchronisation of neuronal responses in the visual cortex induced by stimuli but resulting from internal interactions and therefore not time-locked to the temporal structure of the inducing stimulus. It is undisputed that behaviourally relevant information is encoded in graded and event-related rate fluctuations of neuronal responses (e.g. the rate-coded population vectors in the motor cortex [4]). However, it is currently a matter of controversy whether neurons in the central nervous system can distinguish with sufficient temporal precision between synchronised and temporally dispersed synaptic input in order to exploit precise temporal relations among converging synaptic activity for the encoding of information. Studies pertinent to this issue have been performed at many different levels; for example, electrophysiological analysis of synaptic integration in ex vivo preparations, single- and multi-cell recordings in vivo in a variety of species (ranging from insects to primates), noninvasive examinations of electrical activity in human brains, psychophysical analysis of temporal integration in human subjects and computer simulations. I review here experiments that have examined temporal features of neuronal integration at different levels of analysis.

Section snippets

Transmission of temporal stimulus features

Behavioural analyses and electrophysiological recording have long shown that the neuronal networks involved in binaural sound localisation can resolve time delays between the inputs from the two ears in the submillisecond range, implying extremely precise transmission of temporal stimulus features and hence very short time constants for synaptic integration [5]. Evidence that other systems, including neocortical circuits, are also capable of handling temporal information with a precision in the

Cellular signals

If neuronal networks can operate with high temporal precision and if precisely synchronised synaptic inputs are more effective than temporally dispersed EPSPs, precise temporal relations among the discharges of distributed neurons can, in principle, be exploited for coding. First, synchronisation could be used for response selection because it enhances the saliency of discharges with great temporal selectivity. Second, by virtue of the fact that saliency increases simultaneously and selectively

Synchrony and learning

If precise timing relations among the discharges of distributed neurons are used not only to convey information about the temporal structure of stimuli but also as a mechanism for the selection and grouping of responses, learning mechanisms must operate with similar temporal resolution. Recent evidence indicates that use-dependent modifications of synaptic gain such as long-term potentiation (LTP) and long-term depression (LTD) are indeed dependent on very precise temporal relations between

Synchronisation as a signature of relatedness

The results reviewed so far provide correlative evidence for a role of response synchronisation in neuronal processing but permit no stringent inferences as to whether the nervous system attributes significance to the precise temporal relations among discharges. Recent psychophysical studies in humans, however, have exploited the fact that responses of neurons in the visual cortex can be made synchronous by synchronising the stimuli that evoke them, with a precision similar to that reached by

Conclusions

The studies reviewed here allow us to draw four main conclusions about neuronal networks. First, they can handle temporal information with high precision. Second, they transmit synchronised activity with more efficiency than asynchronous activity. Third, through internal interactions, they can generate precise temporal relations among distributed discharges that are independent of the temporal structure of stimuli. Finally, they interpret the synchronisation of discharges as signature of

References and recommended reading

Papers of particular interest, published within the annual period of review, have been highlighted as:

  • • of special interest

  • •• of outstanding interest

References (70)

  • C.E. Carr

    Processing of temporal information in the brain

    Annu Rev Neurosci

    (1993)
  • R.C. DeCharms et al.

    Optimizing sound features for cortical neurons

    Science

    (1998)
  • M.P. Kilgard et al.

    Plasticity of temporal information processing in the primary auditory cortex

    Nat Neurosci

    (1998)
  • A.C. Yu et al.

    Temporal hierarchical control of singing in birds

    Science

    (1996)
  • A.J. Doupe

    Song- and order-selective neurons in the songbird anterior forebrain and their emergence during vocal development

    J Neurosci

    (1997)
  • G. Rager et al.

    The response of cat visual cortex to flicker stimuli of variable frequency

    Eur J Neurosci

    (1998)
  • M. Castelo-Branco et al.

    Synchronization of visual responses between the cortex, lateral geniculate nucleus, and retina in the anesthetized cat

    J Neurosci

    (1998)
  • S. Neuenschwander et al.

    Long-range synchronization of oscillatory light responses in the cat retina and lateral geniculate nucleus

    Nature

    (1996)
  • S. Herculano et al.

    Precisely synchronized oscillatory firing patterns require EEG activation

    J Neurosci

    (1999)
  • S. Friedman-Hill et al.

    Temporal dynamics of neuronal activity in the striate cortex of alert macaque: I. Incidence and stimulus-dependence of oscillations

    J Neurosci

    (1999)
  • P.E. Maldonado et al.

    Temporal dynamics of neuronal activity in the striate cortex of alert macaque: II. Short and long-range temporally-correlated activity

    J Neurosci

    (1999)
  • M.S. Livingstone

    Oscillatory firing and interneuronal correlations in squirrel monkey striate cortex

    J Neurophysiol

    (1996)
  • M. Brecht et al.

    Correlation analysis of corticotectal interactions in the cat visual system

    J Neurophysiol

    (1998)
  • E.D. Gershon et al.

    Coding strategies in monkey V1 and inferior temporal cortices

    J Neurophysiol

    (1998)
  • R. Lestienne et al.

    The significance of precisely replicating patterns in mammalian CNS spike trains

    Neuroscience

    (1998)
  • Y. Prut et al.

    Spatiotemporal structure of cortical activity: properties and behavioral relevance

    J Neurophysiol

    (1998)
  • M.N. Shadlen et al.

    The variable discharge of cortical neurons: implications for connectivity, computation, and information coding

    J Neurosci

    (1998)
  • M. Abeles

    Corticonics: Neural Circuits of Cerebral Cortex

    (1991)
  • M. Diesmann et al.

    Cortical synfire activity — a two dimensional state space analysis

  • E.D. Lumer et al.

    Neural dynamic in a model of the thalamocortical system. I. Layers, loops and the emergence of fast synchronous rhythms

    Cereb Cortex

    (1997)
  • E.D. Lumer et al.

    Neural dynamics in a model of the thalamocortical system. II. The role of neural synchrony tested through perturbations of spike timing

    Cereb Cortex

    (1997)
  • C.F. Stevens et al.

    Input synchrony and the irregular firing of cortical neurons

    Nat Neurosci

    (1998)
  • M. Matsumura et al.

    Synaptic interactions between primate precentral cortex neurons revealed by spike-triggered averaging of intracellular membrane potentials in vivo

    J Neurosci

    (1996)
  • J.-M. Alonso et al.

    Precisely correlated firing in cells of the lateral geniculate nucleus

    Nature

    (1996)
  • J.-M. Alonso et al.

    Functional connectivity between simple cells and complex cells in cat striate cortex

    Nat Neurosci

    (1998)
  • Cited by (129)

    • Methods for analysis of brain connectivity: An IFCN-sponsored review

      2019, Clinical Neurophysiology
      Citation Excerpt :

      Similarly, network dynamics which do not require changes in energy consumption (i.e. phase locking-unlocking with a stable firing frequency) do not produce a BOLD signal in fMRI. In fact, ‘coding by synchrony’ is possible in the absence of changes in energy consumption and of significant net changes of averaged neuronal population activity over time (Singer, 1999b). This implies that selecting a priori networks on the basis of significantly enhanced local activation may miss relevant network nodes.

    • Effect of spike-timing-dependent plasticity on neural assembly computing

      2016, Neurocomputing
      Citation Excerpt :

      The neural structure of basal ganglia comprises several neural groups, which employ a mechanism similar to neural assemblies [10]. Furthermore, various terms are used to describe neural assemblies in neuroscience, such as persistent activity [11], neural synchrony [12], neural cell assemblies [12,13], spatiotemporal firing patterns [14], and oscillating networks, where assemblies of neurons have functional relations defined by synchronized oscillations, which is referred to as “binding by synchrony” [15]. This model has gained substantial support based on experimental symptom states [16–18].

    • Neurocomputation and Coding in the Claustrum. Comparisons with the Pulvinar.

      2014, The Claustrum: Structural, Functional, and Clinical Neuroscience
    View all citing articles on Scopus
    View full text