Information flow and temporal coding in primate pattern vision

Heller, Joshua A.; Hertz, John; Kjær, Troels W.; Richmond, Barry J.

doi:10.1007/bf00961433

Cited by 147 publications

(119 citation statements)

References 18 publications

Supporting

Mentioning

112

Contrasting

Order By: Relevance

“…The latter approach has been used successfully for comparing many different coding schemes including temporal versus rate coding (41), small versus large time bins (42)(43)(44), and labeled line versus population averages (45). The reason this does not work for correlations, however, is that assessing the importance of correlations is fundamentally different from comparing codes: In the latter, information about the responses is thrown away, and in the former, information about the response distribution is thrown away.…”

Section: Discussionmentioning

confidence: 99%

Decoding neuronal spike trains: How important are correlations?

Nirenberg

Latham²

2003

Proc. Natl. Acad. Sci. U.S.A.

198

208

View full text Add to dashboard Cite

It has been known for >30 years that neuronal spike trains exhibit correlations, that is, the occurrence of a spike at one time is not independent of the occurrence of spikes at other times, both within spike trains from single neurons and across spike trains from multiple neurons. The presence of these correlations has led to the proposal that they might form a key element of the neural code. Specifically, they might act as an extra channel for information, carrying messages about events in the outside world that are not carried by other aspects of the spike trains, such as firing rate. Currently, there is no general consensus about whether this proposal applies to real spike trains in the nervous system. This is largely because it has been hard to separate information carried in correlations from that not carried in correlations. Here we propose a framework for performing this separation. Specifically, we derive an information-theoretic cost function that measures how much harder it is to decode neuronal responses when correlations are ignored than when they are taken into account. This cost function can be readily applied to real neuronal data. E ver since Adrian and Zotterman observed that the firing rate of peripheral touch receptors coded for the pressure applied to a patch of skin (1), neuroscientists have been trying to crack the neural code, that is, to understand the relationship between neuronal activity and events in the outside world. For much of that time, the working hypothesis was that information is carried by firing rate. More recently it has been proposed that firing rate is not the whole story: Information might also be carried in spike patterns, both within spike trains from single neurons (2-6) and across spike trains from multiple neurons (7-9).One aspect of this proposal, an aspect that has led to a great deal of debate, is that correlations in spike patterns may be of particular importance (7)(8)(9)(10)12). It has been known for many years that spike trains contain correlations; that is, the presence of a spike at one time is not independent of the presence of spikes at other times. These correlations exist not just within spike trains but across them as well, with the most common example being synchronous spikes across pairs of cells (13)(14)(15)(16). What has led to the debate is the suggestion that these correlations might form a key aspect of the code. The idea is that they might serve as an extra information channel, conveying messages not carried elsewhere in the spike trains.How might the correlations do this? An example, using synchronous spikes, is shown in Fig. 1a. In this example there are two stimuli, A and B, and two neurons. When stimulus A is presented, the two neurons produce five spikes on average. When stimulus B is presented, they also produce five spikes on average. What is different, though, is the correlational structure of the responses: When stimulus A is presented, the two neurons tend to produce few synchronous spikes, whereas when stimulus B is presented, they ...

show abstract

Section: Discussionmentioning

confidence: 99%

Decoding neuronal spike trains: How important are correlations?

Nirenberg

Latham²

2003

Proc. Natl. Acad. Sci. U.S.A.

198

208

View full text Add to dashboard Cite

show abstract

“…3, c), although we note that the physical stimulus contains much higher temporal frequencies. For transiently presented, static stimuli, this time scale corresponds to the stimulus duration (McClurkin et al, 1991;Heller et al, 1995;Victor and Purpura, 1996). A limiting case for this distinction, not addressed in our study, is to compare timevarying stimuli with constant stimuli Buracas et al, 1998;Warzecha et al, 1998).…”

Section: Temporal Precision Of Firingmentioning

confidence: 99%

Temporal Coding of Visual Information in the Thalamus

2000

View full text Add to dashboard Cite

The amount of information a sensory neuron carries about a stimulus is directly related to response reliability. We recorded from individual neurons in the cat lateral geniculate nucleus (LGN) while presenting randomly modulated visual stimuli. The responses to repeated stimuli were reproducible, whereas the responses evoked by nonrepeated stimuli drawn from the same ensemble were variable. Stimulus-dependent information was quantified directly from the difference in entropy of these neural responses. We show that a single LGN cell can encode much more visual information than had been demonstrated previously, ranging from 15 to 102 bits/sec across our sample of cells. Information rate was correlated with the firing rate of the cell, for a consistent rate of 3.6 Ϯ 0.6 bits/spike (mean Ϯ SD). This information can primarily be attributed to the high temporal precision with which firing probability is modulated; many individual spikes were timed with better than 1 msec precision. We introduce a way to estimate the amount of information encoded in temporal patterns of firing, as distinct from the information in the time varying firing rate at any temporal resolution. Using this method, we find that temporal patterns sometimes introduce redundancy but often encode visual information. The contribution of temporal patterns ranged from Ϫ3.4 to ϩ25.5 bits/sec or from Ϫ9.4 to ϩ24.9% of the total information content of the responses.Key words:LGN; neural coding; information theory; entropy; white noise; reliability; variability Cells in the lateral geniculate nucleus of the thalamus (LGN) respond to spatial and temporal changes in light intensity within their receptive fields. The collective responses of many such cells constitute the input to visual cortex. All stimulus discrimination at the perceptual level must ultimately be supported by reliable differences in the neural response at the level of the LGN cell population. We are therefore interested in measuring the statistical discriminability of LGN responses elicited by different visual stimuli.It has been shown that the LGN can respond to visual stimuli with remarkable temporal precision (Reich et al., 1997). This implies that LGN neurons have the capability to signal information at high rates. Previous estimates of the information in LGN responses have used two general approaches. The first approach, stimulus reconstruction, relies on an explicit model of what the neuron is encoding, as well as an algorithm for decoding it (Bialek et al., 1991;Rieke et al., 1997). This method has been used to place lower bounds on the information encoded by single neurons (Reinagel et al., 1999) or pairs of neurons (Dan et al., 1998) in the LGN in response to dynamic visual stimuli.The second approach, the "direct" method, relies instead on statistical properties of the responses to different stimuli (the entropy of the responses). Because this involves only comparisons of spike trains, without reference to stimulus parameters, we need not know what features of the stimulus the cell ...

show abstract

“…The existence of such early feedback is supported by monkey electrophysiology, which presents evidence for an early modulation of responses in low-order processing during V5 inactivation (Hupe et al, 1998) and also shows that feedforward and feedback axons conduct 10 times faster than local horizontal axons . To achieve functional significance, modulation must happen within about 100 ms from response onset, since most information is transmitted early (Heller et al, 1995;Tovee et al, 1993). Human data show that information is processed very rapidly in the visual cortex.…”

Section: Introductionmentioning

confidence: 99%