Limits...
Decomposing neural synchrony: toward an explanation for near-zero phase-lag in cortical oscillatory networks.

Rajagovindan R, Ding M - PLoS ONE (2008)

Bottom Line: Recent theories have considered the importance of this phenomenon in establishing an effective communication framework among neuronal ensembles.Two factors, among possibly others, can be hypothesized to contribute to the near-zero phase-lag relationship: (1) positively correlated common input with no significant relative time delay and (2) bidirectional interaction.A rigorous mathematical theorem is used for the first time to empirically examine the factors that contribute to this phenomenon.

View Article: PubMed Central - PubMed

Affiliation: The J. Crayton Pruitt Family Department of Biomedical Engineering, University of Florida, Gainesville, Florida, USA.

ABSTRACT

Background: Synchronized oscillation in cortical networks has been suggested as a mechanism for diverse functions ranging from perceptual binding to memory formation to sensorimotor integration. Concomitant with synchronization is the occurrence of near-zero phase-lag often observed between network components. Recent theories have considered the importance of this phenomenon in establishing an effective communication framework among neuronal ensembles.

Methodology/principal findings: Two factors, among possibly others, can be hypothesized to contribute to the near-zero phase-lag relationship: (1) positively correlated common input with no significant relative time delay and (2) bidirectional interaction. Thus far, no empirical test of these hypotheses has been possible for lack of means to tease apart the specific causes underlying the observed synchrony. In this work simulation examples were first used to illustrate the ideas. A quantitative method that decomposes the statistical interdependence between two cortical areas into a feed-forward, a feed-back and a common-input component was then introduced and applied to test the hypotheses on multichannel local field potential recordings from two behaving monkeys.

Conclusion/significance: The near-zero phase-lag phenomenon is important in the study of large-scale oscillatory networks. A rigorous mathematical theorem is used for the first time to empirically examine the factors that contribute to this phenomenon. Given the critical role that oscillatory activity is likely to play in the regulation of biological processes at all levels, the significance of the proposed method may extend beyond systems neuroscience, the level at which the present analysis is conceived and performed.

Show MeSH
Schematic Granger causality graph.(A) monkey GE and (B) monkey LU. Solid arrows indicate directions of causal influence in the beta frequency band that were significant at p<0.005.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2573956&req=5

pone-0003649-g002: Schematic Granger causality graph.(A) monkey GE and (B) monkey LU. Solid arrows indicate directions of causal influence in the beta frequency band that were significant at p<0.005.

Mentions: Granger causality spectra are shown in Fig. 1 for a pair of sites experiencing unidirectional interaction (A) and another pair of sites undergoing bidirectional interaction (B). Figure 2 shows the interaction patterns among the three recording sites in both monkeys in the beta frequency range where the same significance threshold criterion described in [11] were used. Except for (S1,7b) in monkey GE and (M1,S1) in monkey LU, the remaining site pairs in both monkeys exhibited unidirectional interaction. Unlike the other five site pairs where the interaction is concentrated in the beta range (see Fig. 1), the (M1,S1) pair in monkey LU also exhibited significant interaction in the gamma frequency range, in addition to that in the beta range. For this pair, the causal influences in the time-domain where instantaneous causality is most readily interpreted, are confounded and is thus excluded from further analysis. Functionally, the observation that S1 and 7b play a pivotal role in the organization of the network has led to the hypothesis that the beta network supports the maintenance of lever depression by facilitating sensorimotor integration [11], [26], [30].


Decomposing neural synchrony: toward an explanation for near-zero phase-lag in cortical oscillatory networks.

Rajagovindan R, Ding M - PLoS ONE (2008)

Schematic Granger causality graph.(A) monkey GE and (B) monkey LU. Solid arrows indicate directions of causal influence in the beta frequency band that were significant at p<0.005.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2573956&req=5

pone-0003649-g002: Schematic Granger causality graph.(A) monkey GE and (B) monkey LU. Solid arrows indicate directions of causal influence in the beta frequency band that were significant at p<0.005.
Mentions: Granger causality spectra are shown in Fig. 1 for a pair of sites experiencing unidirectional interaction (A) and another pair of sites undergoing bidirectional interaction (B). Figure 2 shows the interaction patterns among the three recording sites in both monkeys in the beta frequency range where the same significance threshold criterion described in [11] were used. Except for (S1,7b) in monkey GE and (M1,S1) in monkey LU, the remaining site pairs in both monkeys exhibited unidirectional interaction. Unlike the other five site pairs where the interaction is concentrated in the beta range (see Fig. 1), the (M1,S1) pair in monkey LU also exhibited significant interaction in the gamma frequency range, in addition to that in the beta range. For this pair, the causal influences in the time-domain where instantaneous causality is most readily interpreted, are confounded and is thus excluded from further analysis. Functionally, the observation that S1 and 7b play a pivotal role in the organization of the network has led to the hypothesis that the beta network supports the maintenance of lever depression by facilitating sensorimotor integration [11], [26], [30].

Bottom Line: Recent theories have considered the importance of this phenomenon in establishing an effective communication framework among neuronal ensembles.Two factors, among possibly others, can be hypothesized to contribute to the near-zero phase-lag relationship: (1) positively correlated common input with no significant relative time delay and (2) bidirectional interaction.A rigorous mathematical theorem is used for the first time to empirically examine the factors that contribute to this phenomenon.

View Article: PubMed Central - PubMed

Affiliation: The J. Crayton Pruitt Family Department of Biomedical Engineering, University of Florida, Gainesville, Florida, USA.

ABSTRACT

Background: Synchronized oscillation in cortical networks has been suggested as a mechanism for diverse functions ranging from perceptual binding to memory formation to sensorimotor integration. Concomitant with synchronization is the occurrence of near-zero phase-lag often observed between network components. Recent theories have considered the importance of this phenomenon in establishing an effective communication framework among neuronal ensembles.

Methodology/principal findings: Two factors, among possibly others, can be hypothesized to contribute to the near-zero phase-lag relationship: (1) positively correlated common input with no significant relative time delay and (2) bidirectional interaction. Thus far, no empirical test of these hypotheses has been possible for lack of means to tease apart the specific causes underlying the observed synchrony. In this work simulation examples were first used to illustrate the ideas. A quantitative method that decomposes the statistical interdependence between two cortical areas into a feed-forward, a feed-back and a common-input component was then introduced and applied to test the hypotheses on multichannel local field potential recordings from two behaving monkeys.

Conclusion/significance: The near-zero phase-lag phenomenon is important in the study of large-scale oscillatory networks. A rigorous mathematical theorem is used for the first time to empirically examine the factors that contribute to this phenomenon. Given the critical role that oscillatory activity is likely to play in the regulation of biological processes at all levels, the significance of the proposed method may extend beyond systems neuroscience, the level at which the present analysis is conceived and performed.

Show MeSH