Information theory and neural coding

Borst, Alexander; Theunissen, Frédéric E.

doi:10.1038/14731

Cited by 990 publications

(914 citation statements)

References 47 publications

Supporting

Mentioning

869

Contrasting

Unclassified

Order By: Relevance

“…distribution function of quantities characterizing stimuli) or about the individual stimuli that caused the spike train under consideration (Borst and Theunissen, 1999)? This is something that, in general, can only be answered out of experience, although the very definition of the coding might help to figure out the relevant aspects involved.…”

Section: Discussionmentioning

confidence: 99%

“…Basically all the information that travels without decay along neuronal axons and is transmitted between neurons, does it in the form of individual spikes or trains of spikes. It is still a challenging problem to find out what kind of encoding and decoding mechanisms are used by the neurons and which are the properties of these biological encoders (Borst and Theunissen, 1999). Important for the mathematical modeler is the fact that the relation between stimulus and neuron response is not one-to-one: the same input s can generate different outputs r k because of noise in the neuron.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

On the number of states of the neuronal sources

et al. 2003

View full text Add to dashboard Cite

In a previous paper (Proceedings of the World Congress on Neuroinformatics (2001)) the authors applied the socalled Lempel Á/Ziv complexity to study neural discharges (spike trains) from an information-theoretical point of view. Along with other results, it is shown there that this concept of complexity allows to characterize the responses of primary visual cortical neurons to both random and periodic stimuli. To this aim we modeled the neurons as information sources and the spike trains as messages generated by them. In this paper, we study further consequences of this mathematical approach, this time concerning the number of states of such neuronal information sources. In this context, the state of an information source means an internal degree of freedom (or parameter) which allows outputs with more general stochastic properties, since symbol generation probabilities at every time step may additionally depend on the value of the current state of the neuron. Furthermore, if the source is ergodic and Markovian, the number of states is directly related to the stochastic dependence lag of the source and provides a measure of the autocorrelation of its messages. Here, we find that the number of states of the neurons depends on the kind of stimulus and the type of preparation ( in vivo versus in vitro recordings), thus providing another way of differentiating neuronal responses. In particular, we observed that (for the encoding methods considered) in vitro sources have a higher lag than in vivo sources for periodic stimuli. This supports the conclusion put forward in the paper mentioned above that, for the same kind of stimulus, in vivo responses are more random (hence, more difficult to compress) than in vitro responses and, consequently, the former transmit more information than the latter. #

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

On the number of states of the neuronal sources

et al. 2003

View full text Add to dashboard Cite

show abstract

“…There are several ways of determining estimates for the mutual information rate (Borst and Theunissen, 1999). We first describe the indirect method, that aims at estimating the mutual information indirectly through the stimulus reconstruction technique.…”

Section: Information Theory and Stimulus Estimationmentioning

confidence: 99%

“…Alternatively, one can use the indirect method to obtain a lower bound on information transfer since this measure relies on linear decoding (Gabbiani, 1996;Rieke et al, 1996). One can thus use the direct method to gauge the goodness of the lower bound estimate obtained with the indirect method (Borst and Theunissen, 1999). A difference between the lower bound and the direct method indicates that some features of the stimulus might be decoded in a non-linear manner.…”

Section: Information Theory: Linear Versus Nonlinear Codingmentioning

confidence: 99%

“…We also use information theory (Borst and Theunissen, 1999) to quantify the amount of information transmitted by each neuron. This information can be computed in two ways: the direct method of computing the mutual information makes no assumptions on the nature of the neural code while the indirect method assumes a linear decoder (Rieke et al, 1996;Borst and Theunissen, 1999) and thus computes a lower bound to the rate of information transmission (Theunissen et al, 1996). We compare results obtained using both methods to estimate the amount of information that can be decoded through both linear and non-linear means.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

To Burst or Not to Burst?

2004

View full text Add to dashboard Cite

It is well known that some neurons tend to fire packets of action potentials followed by periods of quiescence (bursts) while others within the same stage of sensory processing fire in a tonic manner. However, the respective computational advantages of bursting and tonic neurons for encoding time varying signals largely remain a mystery. Weakly electric fish use cutaneous electroreceptors to convey information about sensory stimuli and it has been shown that some electroreceptors exhibit bursting dynamics while others do not. In this study, we compare the neural coding capabilities of tonically firing and bursting electroreceptor model neurons using information theoretic measures. We find that both bursting and tonically firing model neurons efficiently transmit information about the stimulus. However, the decoding mechanisms that must be used for each differ greatly: a non-linear decoder would be required to extract all the available information transmitted by the bursting model neuron whereas a linear one might suffice for the tonically firing model neuron. Further investigations using stimulus reconstruction techniques reveal that, unlike the tonically firing model neuron, the bursting model neuron does not encode the detailed time course of the stimulus. A novel measure of feature detection reveals that the bursting neuron signals certain stimulus features. Finally, we show that feature extraction and stimulus estimation are mutually exclusive computations occurring in bursting and tonically firing model neurons, respectively. Our results therefore suggest that stimulus estimation and feature extraction might be parallel computations in certain sensory systems rather than being sequential as has been previously proposed.

show abstract

Neural Signal Manager: a collection of classical and innovative tools for multi‐channel spike train analysis

Novellino¹,

Chiappalone

Maccione³

et al. 2008

Adaptive Control & Signal

View full text Add to dashboard Cite

SUMMARYRecent developments in the neuroengineering field and the widespread use of the micro electrode arrays (MEAs) for electrophysiological investigations made available new approaches for studying the dynamics of dissociated neuronal networks as well as acute/organotypic slices maintained ex vivo. Importantly, the extraction of relevant parameters from these neural populations is likely to involve long-term measurements, lasting from a few hours to entire days. The processing of huge amounts of electrophysiological data, in terms of computational time and automation of the procedures, is actually one of the major bottlenecks for both in vivo and in vitro recordings. In this paper we present a collection of algorithms implemented within a new software package, named the Neural Signal Manager (NSM), aimed at analyzing a huge quantity of data recorded by means of MEAs in a fast and efficient way. The NSM offers different approaches for both spike and burst analysis, and integrates state-of-the-art statistical algorithms, such as the interspike interval histogram or the post stimulus time histogram, with some recent ones, such as the burst detection and its related statistics. In order to show the potentialities of the software, the application of the developed algorithms to a set of spontaneous activity recordings from dissociated cultures at different ages is presented in the Results section.

show abstract

Information theory and neural coding

Cited by 990 publications

References 47 publications

On the number of states of the neuronal sources

On the number of states of the neuronal sources

To Burst or Not to Burst?

Neural Signal Manager: a collection of classical and innovative tools for multi‐channel spike train analysis

Contact Info

Product

Resources

About