Limits...
Quantifying neural coding of event timing.

Soteropoulos DS, Baker SN - J. Neurophysiol. (2008)

Bottom Line: The perievent time histogram (PETH) examines how, on average, neural firing modulates before and after the alignment event.This is used to generate a probability distribution of the event occurrence, using Bayes' rule.By an information theoretic approach, this method yields a single value (in bits) that quantifies the reduction in uncertainty regarding the time of an external event following observation of the spike train.

View Article: PubMed Central - PubMed

Affiliation: Institute of Neuroscience, Medical School, Newcastle University, Newcastle upon Tyne, NE2 4HH, UK.

ABSTRACT
Single-neuron firing is often analyzed relative to an external event, such as successful task performance or the delivery of a stimulus. The perievent time histogram (PETH) examines how, on average, neural firing modulates before and after the alignment event. However, the PETH contains no information about the single-trial reliability of the neural response, which is important from the perspective of a target neuron. In this study, we propose the concept of using the neural activity to predict the timing of the occurrence of an event, as opposed to using the event to predict the neural response. We first estimate the likelihood of an observed spike train, under the assumption that it was generated by an inhomogeneous gamma process with rate profile similar to the PETH shifted by a small time. This is used to generate a probability distribution of the event occurrence, using Bayes' rule. By an information theoretic approach, this method yields a single value (in bits) that quantifies the reduction in uncertainty regarding the time of an external event following observation of the spike train. We show that the approach is sensitive to the amplitude of a response, to the level of baseline firing, and to the consistency of a response between trials, all of which are factors that will influence a neuron's ability to code for the time of the event. The technique can provide a useful means not only of determining which of several behavioral events a cell encodes best, but also of permitting objective comparison of different cell populations.

Show MeSH
Effect of smoothing, order, trial number, and rate on bias. A: effect of smoothing on PETHs (with no rate step), measured at 3 different baseline firing rates, for 10 trials per simulation. The horizontal dotted line represents the maximum of the ordinate axis in B. A Gaussian kernel of unit area was used for smoothing with different SDs, shown on the abscissa. Each point is the average of 10 simulations. B: like A but 50 trials were used. The dotted line represents the size of the ordinate axis in C. C: like B but 100 trials were used. D: effect of noise, quantified by coefficient of variation (CV) on the bias, for 4 different basal rates. E: effect of spiking regularity on bias for different baseline rates. Ten trials per PETH used. F: like E but 50 trials per PETH used. G: like E but 100 trials per PETH used. Key for A applies to A–C and E–G.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2637006&req=5

f7: Effect of smoothing, order, trial number, and rate on bias. A: effect of smoothing on PETHs (with no rate step), measured at 3 different baseline firing rates, for 10 trials per simulation. The horizontal dotted line represents the maximum of the ordinate axis in B. A Gaussian kernel of unit area was used for smoothing with different SDs, shown on the abscissa. Each point is the average of 10 simulations. B: like A but 50 trials were used. The dotted line represents the size of the ordinate axis in C. C: like B but 100 trials were used. D: effect of noise, quantified by coefficient of variation (CV) on the bias, for 4 different basal rates. E: effect of spiking regularity on bias for different baseline rates. Ten trials per PETH used. F: like E but 50 trials per PETH used. G: like E but 100 trials per PETH used. Key for A applies to A–C and E–G.

Mentions: With finite data, the method presented here will lead to a nonzero raw information estimate Ir, even if a cell does not modulate its activity systematically around a task event. Any noise fluctuations in the shift probability estimate will lower its entropy relative to a uniform distribution, leading to a positive bias in the information. We have used a shuffling method to estimate and correct for this. The dependence of the bias on various parameters is illustrated in Fig. 7; these curves were calculated from simulated data with no actual modulation in rate.


Quantifying neural coding of event timing.

Soteropoulos DS, Baker SN - J. Neurophysiol. (2008)

Effect of smoothing, order, trial number, and rate on bias. A: effect of smoothing on PETHs (with no rate step), measured at 3 different baseline firing rates, for 10 trials per simulation. The horizontal dotted line represents the maximum of the ordinate axis in B. A Gaussian kernel of unit area was used for smoothing with different SDs, shown on the abscissa. Each point is the average of 10 simulations. B: like A but 50 trials were used. The dotted line represents the size of the ordinate axis in C. C: like B but 100 trials were used. D: effect of noise, quantified by coefficient of variation (CV) on the bias, for 4 different basal rates. E: effect of spiking regularity on bias for different baseline rates. Ten trials per PETH used. F: like E but 50 trials per PETH used. G: like E but 100 trials per PETH used. Key for A applies to A–C and E–G.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2637006&req=5

f7: Effect of smoothing, order, trial number, and rate on bias. A: effect of smoothing on PETHs (with no rate step), measured at 3 different baseline firing rates, for 10 trials per simulation. The horizontal dotted line represents the maximum of the ordinate axis in B. A Gaussian kernel of unit area was used for smoothing with different SDs, shown on the abscissa. Each point is the average of 10 simulations. B: like A but 50 trials were used. The dotted line represents the size of the ordinate axis in C. C: like B but 100 trials were used. D: effect of noise, quantified by coefficient of variation (CV) on the bias, for 4 different basal rates. E: effect of spiking regularity on bias for different baseline rates. Ten trials per PETH used. F: like E but 50 trials per PETH used. G: like E but 100 trials per PETH used. Key for A applies to A–C and E–G.
Mentions: With finite data, the method presented here will lead to a nonzero raw information estimate Ir, even if a cell does not modulate its activity systematically around a task event. Any noise fluctuations in the shift probability estimate will lower its entropy relative to a uniform distribution, leading to a positive bias in the information. We have used a shuffling method to estimate and correct for this. The dependence of the bias on various parameters is illustrated in Fig. 7; these curves were calculated from simulated data with no actual modulation in rate.

Bottom Line: The perievent time histogram (PETH) examines how, on average, neural firing modulates before and after the alignment event.This is used to generate a probability distribution of the event occurrence, using Bayes' rule.By an information theoretic approach, this method yields a single value (in bits) that quantifies the reduction in uncertainty regarding the time of an external event following observation of the spike train.

View Article: PubMed Central - PubMed

Affiliation: Institute of Neuroscience, Medical School, Newcastle University, Newcastle upon Tyne, NE2 4HH, UK.

ABSTRACT
Single-neuron firing is often analyzed relative to an external event, such as successful task performance or the delivery of a stimulus. The perievent time histogram (PETH) examines how, on average, neural firing modulates before and after the alignment event. However, the PETH contains no information about the single-trial reliability of the neural response, which is important from the perspective of a target neuron. In this study, we propose the concept of using the neural activity to predict the timing of the occurrence of an event, as opposed to using the event to predict the neural response. We first estimate the likelihood of an observed spike train, under the assumption that it was generated by an inhomogeneous gamma process with rate profile similar to the PETH shifted by a small time. This is used to generate a probability distribution of the event occurrence, using Bayes' rule. By an information theoretic approach, this method yields a single value (in bits) that quantifies the reduction in uncertainty regarding the time of an external event following observation of the spike train. We show that the approach is sensitive to the amplitude of a response, to the level of baseline firing, and to the consistency of a response between trials, all of which are factors that will influence a neuron's ability to code for the time of the event. The technique can provide a useful means not only of determining which of several behavioral events a cell encodes best, but also of permitting objective comparison of different cell populations.

Show MeSH