Limits...
High baseline activity in inferior temporal cortex improves neural and behavioral discriminability during visual categorization.

Emadi N, Rajimehr R, Esteky H - Front Syst Neurosci (2014)

Bottom Line: However, it is not known how the baseline activity contributes to neural coding and behavior.Specifically we found that a low-frequency (<8 Hz) oscillation in the spike train, prior and phase-locked to the stimulus onset, was correlated with increased gamma power and neuronal baseline activity.This enhancement of the baseline activity was then followed by an increase in the neural selectivity and the response reliability and eventually a higher behavioral performance.

View Article: PubMed Central - PubMed

Affiliation: School of Cognitive Sciences, Institute for Research in Fundamental Sciences (IPM) Tehran, Iran ; Research Center for Brain and Cognition, School of Medicine, University of Shahid Beheshti Tehran, Iran ; Howard Hughes Medical Institute and Department of Neurobiology, Stanford University School of Medicine Stanford, CA, USA.

ABSTRACT
Spontaneous firing is a ubiquitous property of neural activity in the brain. Recent literature suggests that this baseline activity plays a key role in perception. However, it is not known how the baseline activity contributes to neural coding and behavior. Here, by recording from the single neurons in the inferior temporal cortex of monkeys performing a visual categorization task, we thoroughly explored the relationship between baseline activity, the evoked response, and behavior. Specifically we found that a low-frequency (<8 Hz) oscillation in the spike train, prior and phase-locked to the stimulus onset, was correlated with increased gamma power and neuronal baseline activity. This enhancement of the baseline activity was then followed by an increase in the neural selectivity and the response reliability and eventually a higher behavioral performance.

No MeSH data available.


Neural events following baseline modulation during a categorization task. (A) Schematic diagram of the mean response of a model body neuron responding to body and non-body images in “correct” trials. The impact of rhythmic baseline modulation on the neural response and behavior is illustrated. X-axis as in (B). (B) Plot of normalized averaged firing rate of body neurons in correct trials. In each neuron and each signal level, the firing rates were normalized by the peak response, and then the normalized firing rates were averaged. (C) Schematic diagram of the mean response of a model body neuron responding to body and non-body images in “wrong” trials. The impact of no baseline activity on the neural response and behavior is illustrated. X-axis as in (D). (D) Plot of normalized averaged firing rate of body neurons in wrong trials. Normalization was done similar to (B).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4217374&req=5

Figure 9: Neural events following baseline modulation during a categorization task. (A) Schematic diagram of the mean response of a model body neuron responding to body and non-body images in “correct” trials. The impact of rhythmic baseline modulation on the neural response and behavior is illustrated. X-axis as in (B). (B) Plot of normalized averaged firing rate of body neurons in correct trials. In each neuron and each signal level, the firing rates were normalized by the peak response, and then the normalized firing rates were averaged. (C) Schematic diagram of the mean response of a model body neuron responding to body and non-body images in “wrong” trials. The impact of no baseline activity on the neural response and behavior is illustrated. X-axis as in (D). (D) Plot of normalized averaged firing rate of body neurons in wrong trials. Normalization was done similar to (B).

Mentions: Using a schematic model, we have summarized our findings to describe how a correct choice is made during the categorization task (Figure 9). A synchronous oscillation of baseline activity occurs across a population of IT neurons. The strength of such synchronous oscillatory activity can vary across trials, based on the level of cognitive factors such as attention and motivation. In trials with a strong low-frequency (<8 Hz) oscillation, that is coupled with the gamma band and phase-locked to the stimulus onset, a peak of the oscillation can effectively occur before the stimulus presentation; a situation that results in an apparent “baseline shift.” Another peak of the oscillation can also occur around the stimulus presentation, which would enhance the neural responsiveness and produce an elevated evoked response. The enhanced baseline and evoked activity in these oscillatory/HBT trials subsequently increases neural selectivity and reduces response variability (two signatures of an improved neural performance (Treue and Martinez Trujillo, 1999; Mitchell et al., 2007). A higher neural performance would eventually lead to an increased probability of correct choices (Figure 9A). Consistent with this model, the response of body neurons in correct trials of our experiment shows a rhythmic baseline shift, a higher response selectivity and reliability, and subsequently a larger difference between the evoked responses to the preferred and non-preferred categories (Figure 9B). On the other hand, in trials lacking the synchronized oscillatory activity, the baseline activity is constantly low. Stimuli presented in this state evoke low-amplitude responses, with less selectivity and reliability, followed by wrong choices (Figure 9C). The response of body neurons in wrong trials of our experiment is consistent with this model (Figure 9D). The model predicts that, for decision making about the stimulus, a decision boundary could be set efficiently in HBT as a result of higher response discriminability and lower response variability. In contrast in LBT, responses to different categories are mixed and no clear decision boundary would exist. Both predictions are consistent with our data.


High baseline activity in inferior temporal cortex improves neural and behavioral discriminability during visual categorization.

Emadi N, Rajimehr R, Esteky H - Front Syst Neurosci (2014)

Neural events following baseline modulation during a categorization task. (A) Schematic diagram of the mean response of a model body neuron responding to body and non-body images in “correct” trials. The impact of rhythmic baseline modulation on the neural response and behavior is illustrated. X-axis as in (B). (B) Plot of normalized averaged firing rate of body neurons in correct trials. In each neuron and each signal level, the firing rates were normalized by the peak response, and then the normalized firing rates were averaged. (C) Schematic diagram of the mean response of a model body neuron responding to body and non-body images in “wrong” trials. The impact of no baseline activity on the neural response and behavior is illustrated. X-axis as in (D). (D) Plot of normalized averaged firing rate of body neurons in wrong trials. Normalization was done similar to (B).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4217374&req=5

Figure 9: Neural events following baseline modulation during a categorization task. (A) Schematic diagram of the mean response of a model body neuron responding to body and non-body images in “correct” trials. The impact of rhythmic baseline modulation on the neural response and behavior is illustrated. X-axis as in (B). (B) Plot of normalized averaged firing rate of body neurons in correct trials. In each neuron and each signal level, the firing rates were normalized by the peak response, and then the normalized firing rates were averaged. (C) Schematic diagram of the mean response of a model body neuron responding to body and non-body images in “wrong” trials. The impact of no baseline activity on the neural response and behavior is illustrated. X-axis as in (D). (D) Plot of normalized averaged firing rate of body neurons in wrong trials. Normalization was done similar to (B).
Mentions: Using a schematic model, we have summarized our findings to describe how a correct choice is made during the categorization task (Figure 9). A synchronous oscillation of baseline activity occurs across a population of IT neurons. The strength of such synchronous oscillatory activity can vary across trials, based on the level of cognitive factors such as attention and motivation. In trials with a strong low-frequency (<8 Hz) oscillation, that is coupled with the gamma band and phase-locked to the stimulus onset, a peak of the oscillation can effectively occur before the stimulus presentation; a situation that results in an apparent “baseline shift.” Another peak of the oscillation can also occur around the stimulus presentation, which would enhance the neural responsiveness and produce an elevated evoked response. The enhanced baseline and evoked activity in these oscillatory/HBT trials subsequently increases neural selectivity and reduces response variability (two signatures of an improved neural performance (Treue and Martinez Trujillo, 1999; Mitchell et al., 2007). A higher neural performance would eventually lead to an increased probability of correct choices (Figure 9A). Consistent with this model, the response of body neurons in correct trials of our experiment shows a rhythmic baseline shift, a higher response selectivity and reliability, and subsequently a larger difference between the evoked responses to the preferred and non-preferred categories (Figure 9B). On the other hand, in trials lacking the synchronized oscillatory activity, the baseline activity is constantly low. Stimuli presented in this state evoke low-amplitude responses, with less selectivity and reliability, followed by wrong choices (Figure 9C). The response of body neurons in wrong trials of our experiment is consistent with this model (Figure 9D). The model predicts that, for decision making about the stimulus, a decision boundary could be set efficiently in HBT as a result of higher response discriminability and lower response variability. In contrast in LBT, responses to different categories are mixed and no clear decision boundary would exist. Both predictions are consistent with our data.

Bottom Line: However, it is not known how the baseline activity contributes to neural coding and behavior.Specifically we found that a low-frequency (<8 Hz) oscillation in the spike train, prior and phase-locked to the stimulus onset, was correlated with increased gamma power and neuronal baseline activity.This enhancement of the baseline activity was then followed by an increase in the neural selectivity and the response reliability and eventually a higher behavioral performance.

View Article: PubMed Central - PubMed

Affiliation: School of Cognitive Sciences, Institute for Research in Fundamental Sciences (IPM) Tehran, Iran ; Research Center for Brain and Cognition, School of Medicine, University of Shahid Beheshti Tehran, Iran ; Howard Hughes Medical Institute and Department of Neurobiology, Stanford University School of Medicine Stanford, CA, USA.

ABSTRACT
Spontaneous firing is a ubiquitous property of neural activity in the brain. Recent literature suggests that this baseline activity plays a key role in perception. However, it is not known how the baseline activity contributes to neural coding and behavior. Here, by recording from the single neurons in the inferior temporal cortex of monkeys performing a visual categorization task, we thoroughly explored the relationship between baseline activity, the evoked response, and behavior. Specifically we found that a low-frequency (<8 Hz) oscillation in the spike train, prior and phase-locked to the stimulus onset, was correlated with increased gamma power and neuronal baseline activity. This enhancement of the baseline activity was then followed by an increase in the neural selectivity and the response reliability and eventually a higher behavioral performance.

No MeSH data available.