Limits...
A micro-pool model for decision-related signals in visual cortical areas.

Parker AJ - Front Comput Neurosci (2013)

Bottom Line: However, recent analyses show that the size of a decision-related change in firing in a particular neuron is not a secure basis for concluding anything about the contribution of a single neuron to the formation of a decision: rather the size of the decision-related firing is expected to be dominated by the extent to which the activation of a single neuron is correlated with the firing of the pool of neurons.This article examines the consequences of such a proposal within the visual nervous system.The main focus is on the signals available from single neurons, but it argued that models of choice-related signals must scale up to larger numbers of neurons because MRI and MEG studies also show evidence of similar choice signals.

View Article: PubMed Central - PubMed

Affiliation: Department of Physiology, Anatomy and Genetics, University of Oxford Oxford, UK.

ABSTRACT
The study of sensory signaling in the visual cortex has been greatly advanced by the recording of neural activity simultaneously with the performance of a specific psychophysical task. Individual nerve cells may also increase their firing leading up to the particular choice or decision made on a single psychophysical trial. Understanding these signals is important because they have been taken as evidence that a particular nerve cell or group of nerve cells in the cortex is involved in the formation of the perceptual decision ultimately signaled by the organism. However, recent analyses show that the size of a decision-related change in firing in a particular neuron is not a secure basis for concluding anything about the contribution of a single neuron to the formation of a decision: rather the size of the decision-related firing is expected to be dominated by the extent to which the activation of a single neuron is correlated with the firing of the pool of neurons. The critical question becomes what defines membership of a population of neurons. This article presents the proposal that groups of neurons are naturally linked together by their connectivity, which in turn reflects the previous history of sensory stimulations. When a new psychophysical task is performed, a group of neurons relevant to the judgment becomes involved because the firing of some neurons in that group is strongly relevant to the task. This group of neurons is called a micro-pool. This article examines the consequences of such a proposal within the visual nervous system. The main focus is on the signals available from single neurons, but it argued that models of choice-related signals must scale up to larger numbers of neurons because MRI and MEG studies also show evidence of similar choice signals.

No MeSH data available.


Illustrating the effects of correlation on neuronal signals in a population. This simulates a network such as that in Figure 1. (A) Four neurons arranged to have an interneuronal correlation (rho) equal to 0.4. Neurons 2 to 4 are simulated as zero mean, unit variance Gaussian random variables. Neuron 1 is the same, except an artificial choice signal (value = 1.5) is inserted. (B) The responses of the four neurons are subjected to ROC analysis to reveal the choice related signal. The black curve projecting up to the upper left corner represents the detectability of the choice signal on Neuron 1. The colored curves are for Neurons 2,3,4. These neurons received no choice signal in themselves but acquired the choice signal by virtue of their correlated variance with neuron 1. (C) Summary histogram of choice probabilities induced in population of 40 neurons by choice signal inserted into single neuron and presence of interneuronal correlations inducing the choice probability in the other 39. Estimate of 100 repeated trials with a choice signal of 1.5 as at Figure 2.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3753562&req=5

Figure 2: Illustrating the effects of correlation on neuronal signals in a population. This simulates a network such as that in Figure 1. (A) Four neurons arranged to have an interneuronal correlation (rho) equal to 0.4. Neurons 2 to 4 are simulated as zero mean, unit variance Gaussian random variables. Neuron 1 is the same, except an artificial choice signal (value = 1.5) is inserted. (B) The responses of the four neurons are subjected to ROC analysis to reveal the choice related signal. The black curve projecting up to the upper left corner represents the detectability of the choice signal on Neuron 1. The colored curves are for Neurons 2,3,4. These neurons received no choice signal in themselves but acquired the choice signal by virtue of their correlated variance with neuron 1. (C) Summary histogram of choice probabilities induced in population of 40 neurons by choice signal inserted into single neuron and presence of interneuronal correlations inducing the choice probability in the other 39. Estimate of 100 repeated trials with a choice signal of 1.5 as at Figure 2.

Mentions: This principle is illustrated in the simulation illustrated in Figure 2. A small set of neurons was simulated with Gaussian random variables. To just one of these neurons, a choice signal was added—in this case by adding an additional level of activity to that neuron, thereby simulating the effect of a single top-down signal targeting just one neuron in the pool. The pool of neurons is interconnected, thereby generating correlated activity in that pool (Bair et al., 2001).


A micro-pool model for decision-related signals in visual cortical areas.

Parker AJ - Front Comput Neurosci (2013)

Illustrating the effects of correlation on neuronal signals in a population. This simulates a network such as that in Figure 1. (A) Four neurons arranged to have an interneuronal correlation (rho) equal to 0.4. Neurons 2 to 4 are simulated as zero mean, unit variance Gaussian random variables. Neuron 1 is the same, except an artificial choice signal (value = 1.5) is inserted. (B) The responses of the four neurons are subjected to ROC analysis to reveal the choice related signal. The black curve projecting up to the upper left corner represents the detectability of the choice signal on Neuron 1. The colored curves are for Neurons 2,3,4. These neurons received no choice signal in themselves but acquired the choice signal by virtue of their correlated variance with neuron 1. (C) Summary histogram of choice probabilities induced in population of 40 neurons by choice signal inserted into single neuron and presence of interneuronal correlations inducing the choice probability in the other 39. Estimate of 100 repeated trials with a choice signal of 1.5 as at Figure 2.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3753562&req=5

Figure 2: Illustrating the effects of correlation on neuronal signals in a population. This simulates a network such as that in Figure 1. (A) Four neurons arranged to have an interneuronal correlation (rho) equal to 0.4. Neurons 2 to 4 are simulated as zero mean, unit variance Gaussian random variables. Neuron 1 is the same, except an artificial choice signal (value = 1.5) is inserted. (B) The responses of the four neurons are subjected to ROC analysis to reveal the choice related signal. The black curve projecting up to the upper left corner represents the detectability of the choice signal on Neuron 1. The colored curves are for Neurons 2,3,4. These neurons received no choice signal in themselves but acquired the choice signal by virtue of their correlated variance with neuron 1. (C) Summary histogram of choice probabilities induced in population of 40 neurons by choice signal inserted into single neuron and presence of interneuronal correlations inducing the choice probability in the other 39. Estimate of 100 repeated trials with a choice signal of 1.5 as at Figure 2.
Mentions: This principle is illustrated in the simulation illustrated in Figure 2. A small set of neurons was simulated with Gaussian random variables. To just one of these neurons, a choice signal was added—in this case by adding an additional level of activity to that neuron, thereby simulating the effect of a single top-down signal targeting just one neuron in the pool. The pool of neurons is interconnected, thereby generating correlated activity in that pool (Bair et al., 2001).

Bottom Line: However, recent analyses show that the size of a decision-related change in firing in a particular neuron is not a secure basis for concluding anything about the contribution of a single neuron to the formation of a decision: rather the size of the decision-related firing is expected to be dominated by the extent to which the activation of a single neuron is correlated with the firing of the pool of neurons.This article examines the consequences of such a proposal within the visual nervous system.The main focus is on the signals available from single neurons, but it argued that models of choice-related signals must scale up to larger numbers of neurons because MRI and MEG studies also show evidence of similar choice signals.

View Article: PubMed Central - PubMed

Affiliation: Department of Physiology, Anatomy and Genetics, University of Oxford Oxford, UK.

ABSTRACT
The study of sensory signaling in the visual cortex has been greatly advanced by the recording of neural activity simultaneously with the performance of a specific psychophysical task. Individual nerve cells may also increase their firing leading up to the particular choice or decision made on a single psychophysical trial. Understanding these signals is important because they have been taken as evidence that a particular nerve cell or group of nerve cells in the cortex is involved in the formation of the perceptual decision ultimately signaled by the organism. However, recent analyses show that the size of a decision-related change in firing in a particular neuron is not a secure basis for concluding anything about the contribution of a single neuron to the formation of a decision: rather the size of the decision-related firing is expected to be dominated by the extent to which the activation of a single neuron is correlated with the firing of the pool of neurons. The critical question becomes what defines membership of a population of neurons. This article presents the proposal that groups of neurons are naturally linked together by their connectivity, which in turn reflects the previous history of sensory stimulations. When a new psychophysical task is performed, a group of neurons relevant to the judgment becomes involved because the firing of some neurons in that group is strongly relevant to the task. This group of neurons is called a micro-pool. This article examines the consequences of such a proposal within the visual nervous system. The main focus is on the signals available from single neurons, but it argued that models of choice-related signals must scale up to larger numbers of neurons because MRI and MEG studies also show evidence of similar choice signals.

No MeSH data available.