From synchrony to sparseness

doi:10.1016/S0166-2236(02)00016-4

Trends in Neurosciences

Volume 26, Issue 2, February 2003, Pages 61-64

https://doi.org/10.1016/S0166-2236(02)00016-4 Get rights and content

Abstract

The neurons in the antennal lobe of the locust had been shown to encode the identity of odorants using spatially distributed synchronized patterns of neural activity. Recent work describes how such neural patterns are detected. By using non-linear membrane properties, one set of target neurons, the Kenyon cells of the mushroom bodies, are able to act as coincidence detectors, sensitive to synchronized activity. In addition, the specific circuitry between the antennal lobe and the mushroom body refines the spatial–temporal selectivity of the Kenyon cells. In this process, the neural representation of odor identity is transformed from a dense spatial–temporal code into a sparse code.

Section snippets

Neural circuits and neural code in the locust olfactory system

The olfactory system of the locust has been a remarkable model system for the study of the neural code (Fig. 1). Odorants are first detected by 90 000 olfactory receptors distributed on each antenna of the animal. Odorant identity is encoded in the olfactory receptor neurons by their mean firing rates, with many neurons responding to any particular odor. This spatially distributed code is then transformed into the spatial–temporal code found in the antennal lobe. The 830 projection neurons of

Intermediate conclusion

Perez-Orive et al. have shown that the properties of the Kenyon cells and the circuitry between the antennal lobe and the mushroom bodies favor the detection of the synchronized spatial–temporal patterns that are generated in the antennal lobe, and that this tuning results in a drastic transformation of the code for odor identity. It would be difficult, therefore, to argue that the synchrony found in the antennal lobe and all of the specialized circuitry apparently designed to detect it is an

Sparse codes

Theoreticians have provided answers to the first question ever since the term sparse-coding was proposed to describe neural code. In one of the earlier formulations, David Field suggested that sparse codes are ultimately related to the statistics of natural stimulus ensembles [17]. It is presently believed that many of the relevant features within natural scenes produce high-order correlations between the points within an image (e.g. curved edges within an image influence three-point

Dense spatial–temporal codes

The second theoretical question is whether the dense spatial–temporal code that was found at the level of the antennal lobe in the locust is a necessary intermediate step to ultimately obtaining the sparse representation found in the mushroom bodies. The answer to that question remains more elusive. One could envisage circuitry that generates a sparse code directly from a spatially distributed representation of sensory information, in which the information is represented by mean rates. For

Conclusion

The dense spatial–temporal neural code found in the AL is a highly efficient combinatorial code that might be needed both for the encoding of a large ensemble of odors with a limited number of neurons and for the formation of short-term olfactory memories. The sparse code found in the MB represents odor identity in a synthetic fashion that could facilitate the formation of long-term associative memories. The circuitry in the olfactory system of the locust exploits the synchronicity found in the

Acknowledgements

I would like to thank W. Vinje, J. Gallant and S. Shaevitz for their helpful comments on the manuscript.

References (26)

W. Bialek et al.
Reliability and information transmission in spiking neurons
Trends Neurosci.
(1992)
J.J. Eggermont
Is there a neural code?
Neurosci. Biobehav. Rev.
(1998)
W. Bair
Spike timing in the mammalian visual system
Curr. Opin. Neurobiol.
(1999)
H. Spors et al.
Spatio-temporal dynamics of odor representations in the mammalian olfactory bulb
Neuron
(2002)
N. Schweighofer
Unsupervised learning of granule cell sparse codes enhances cerebellar adaptive control
Neuroscience
(2001)
T. Zars
Behavioral functions of the insect mushroom bodies
Curr. Opin. Neurobiol.
(2000)
F. Theunissen et al.
Temporal encoding in nervous systems: a rigorous definition
J. Comput. Neurosci.
(1995)
G. Laurent
A systems perspective on early olfactory coding
Science
(1999)
A. Pouget
Information processing with population codes
Nat. Rev. Neurosci.
(2000)
A.K. Engel
Temporal coding in the visual cortex: new vistas on integration in the nervous system
Trends Neurosci.
(1992)

M. Wehr et al.

Odour encoding by temporal sequences of firing in oscillating neural assemblies

Nature

(1996)

G. Laurent

Temporal representations of odors in an olfactory network

J. Neurosci.

(1996)

A.K. Engel

Dynamic predictions: oscillations and synchrony in top-down processing

Nat. Rev. Neurosci.

(2001)

Cited by (30)

Cortical Mechanisms of Smooth Eye Movements Revealed by Dynamic Covariations of Neural and Behavioral Responses
2008, Neuron
Citation Excerpt :
The more diverse pattern of responses of FEFSEM neurons may be critical for highly specific control of movement on the basis of cortical plans. Recent work in motor control in songbirds and representation of visual stimuli provides evidence that ethologically relevant stimuli and behaviors can be quite sparsely represented, with remarkable sensitivity and precision; here, we provide complementary evidence for temporally sparse coding in a similarly relevant behavioral space (Hahnloser et al., 2002; Theunissen, 2003; Vinje and Gallant, 2000). Because smooth pursuit is considerably slower than most behaviors, temporal sparseness in pursuit-related populations of neurons ought be more restricted than in songbird motor nuclei, which drive a less temporally correlated set of behaviors.
Neural activity in the frontal eye fields controls smooth pursuit eye movements, but the relationship between single neuron responses, cortical population responses, and eye movements is not well understood. We describe an approach to dynamically link trial-to-trial fluctuations in neural responses to parallel variations in pursuit and demonstrate that individual neurons predict eye velocity fluctuations at particular moments during the course of behavior, while the population of neurons collectively tiles the entire duration of the movement. The analysis also reveals the strength of correlations in the eye movement predictions derived from pairs of simultaneously recorded neurons and suggests a simple model of cortical processing. These findings constrain the primate cortical code for movement, suggesting that either a few neurons are sufficient to drive pursuit at any given time or that many neurons operate collectively at each moment with remarkably little variation added to motor command signals downstream from the cortex.
Sparse coding in the neocortex
2007, Evolution of Nervous Systems
Sparse coding has been proposed as a guiding principle in neural representations of sensory input, particularly in the visual system. Because sparse codes are defined as representations with low activity ratios – i.e., at any given time a small proportion of neurons active – they are sometimes proposed as a means to help conserve metabolic costs. Although we accept that such metabolic costs play a role in evolutionary selection, we also argue that sparse coding offers a useful solution to the problem of representing natural data because such a scheme allows the system to take advantage of the sparse structure of the sensory environment. Given the highly regular and sparse structure of natural data (e.g., natural scenes), a variety of recent studies suggest that the sparse coding in the cortex is an adaptation that has evolved to offer an efficient strategy for storing and building associations with natural inputs. However, highly selective sparse codes require a large number of neurons and can also pose problems for learning. We argue that larger brains have evolved strategies that allow neurons to be both highly selective and invariant across irrelevant dimensions. We argue that this invariance is what allows such highly sparse codes to be useful.
Sparse coding of sensory inputs
2004, Current Opinion in Neurobiology
Several theoretical, computational, and experimental studies suggest that neurons encode sensory information using a small number of active neurons at any given point in time. This strategy, referred to as ‘sparse coding’, could possibly confer several advantages. First, it allows for increased storage capacity in associative memories; second, it makes the structure in natural signals explicit; third, it represents complex data in a way that is easier to read out at subsequent levels of processing; and fourth, it saves energy. Recent physiological recordings from sensory neurons have indicated that sparse coding could be a ubiquitous strategy employed in several different modalities across different organisms.
Neural correlates of learned song in the avian forebrain: Simultaneous representation of self and others
2004, Current Opinion in Neurobiology
Songbirds are extraordinary vocalists and sensitive listeners, singing to communicate identity, engage other birds in acoustical combat, and attract mates. These processes involve auditory plasticity in that birds rapidly learn to discriminate novel from familiar songs. Songbirds also are one of the few non-human animals that use auditory feedback to learn their vocalizations, thus auditory–vocal interactions are likely to be important to vocal learning. Recent advances strengthen the connection between song recognition and processing of birdsong in the auditory telencephalon. New insights also have emerged into the mechanisms underlying the ‘gating’ of auditory responses and the emergence of highly selective responses, two processes that could facilitate auditory feedback important to song learning.
On the sparse structure of natural sounds and natural images: Similarities, differences, and implications for neural coding
2019, Frontiers in Computational Neuroscience
Increase in mutual information during interaction with the environment contributes to perception
2019, Entropy

View all citing articles on Scopus

View full text

Trends in Neurosciences

Research FocusFrom synchrony to sparseness

Abstract

Section snippets

Neural circuits and neural code in the locust olfactory system

Intermediate conclusion

Sparse codes

Dense spatial–temporal codes

Conclusion

Acknowledgements

Trends Neurosci.

Neurosci. Biobehav. Rev.

Curr. Opin. Neurobiol.

Neuron

Neuroscience

Curr. Opin. Neurobiol.

Temporal encoding in nervous systems: a rigorous definition

J. Comput. Neurosci.

A systems perspective on early olfactory coding

Science

Information processing with population codes

Nat. Rev. Neurosci.

Temporal coding in the visual cortex: new vistas on integration in the nervous system

Trends Neurosci.

Odour encoding by temporal sequences of firing in oscillating neural assemblies

Nature

Temporal representations of odors in an olfactory network

J. Neurosci.

Dynamic predictions: oscillations and synchrony in top-down processing

Nat. Rev. Neurosci.

Research Focus
From synchrony to sparseness