Serial representation of items during working memory maintenance at letter-selective cortical sites

Bahramisharif, Ali; Jensen, Ole; Jacobs, Joshua; Lisman, John

doi:10.1371/journal.pbio.2003805

Cited by 104 publications

(99 citation statements)

References 90 publications

(124 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…PAC has been suggested to reflect the regulation of sensory information processing achieved in higher frequencies through excitability fluctuations imposed by slower oscillations [9,10,34,36,37,42]. These results, despite methodological differences, are similar to previous findings that have reported PAC in the rat hippocampus [46,47], non-human primates [82,83], and in human intracranial EEG [48][49][50][51]60] and EEG and MEG recordings [54][55][56][84][85][86]. In contrast to PAC, inter-areal CFS connected the phase of  oscillations with the phases of and oscillations only with small frequency ratios (1:2 and 1:3).…”

Section: Large-scale Cfc Network Characterize Human Resting-state Brsupporting

confidence: 88%

“…During task performance, PAC has been suggested to reflect the regulation of sensory information processing in  and -frequencies by excitability fluctuations imposed by  and  oscillations [9,10,34,36,37,42]. A large number of studies have identified local PAC, i.e., PAC observed between different frequency bands of the same signal, between the phases of slower oscillations in delta (13 Hz, theta (37Hz) and  frequency bands and the amplitude of  oscillations in local field potentials (LFPs) in rats [43][44][45][46][47] and in human intracranial EEG [48][49][50][51] and MEG data [52][53][54][55][56]. Such local PAC has cortex-wide spatial modes akin to RSNs [53].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Genuine cross-frequency coupling networks in human resting-state electrophysiological recordings

Siebenhuehner

Wang

Arnulfo

et al. 2019

Preprint

View full text Add to dashboard Cite

AbstractPhase synchronization of neuronal oscillations in specific frequency bands coordinates anatomically distributed neuronal processing and communication. Typically, oscillations and synchronization take place concurrently in many distinct frequencies, which serve separate computational roles in cognitive functions. While within-frequency phase synchronization has been studied extensively, less is known about the mechanisms that govern neuronal processing distributed across frequencies and brain regions. Such integration of processing between frequencies could be achieved via cross-frequency coupling (CFC), either by phase-amplitude coupling (PAC) or by n:m-cross-frequency phase synchrony (CFS). So far, studies have mostly focused on local CFC in individual brain regions, whereas the presence and functional organization of CFC between brain areas have remained largely unknown. We posit that inter-areal CFC may be essential for large-scale coordination of neuronal activity and investigate here whether genuine CFC networks are present in human resting-state brain activity.To assess the functional organization of CFC networks, we identified brain-wide CFC networks at meso-scale resolution from stereo-electroencephalography (SEEG) and at macro-scale resolution from source-reconstructed magnetoencephalography (MEG) data. We developed a novel graph-theoretical method to distinguish genuine CFC from spurious CFC that may arise from non-sinusoidal signals ubiquitous in neuronal activity. We show that genuine inter-areal CFC is present in human resting-state activity in both MEG and SEEG data. Both CFS and PAC networks coupled theta and alpha oscillations with higher frequencies in large-scale networks connecting anterior and posterior brain regions. CFS and PAC networks had distinct spectral patterns and opposing distribution of low- and high frequency network hubs, implying that they constitute distinct CFC mechanisms. The strength of CFS networks was also predictive of cognitive performance in a separate neuropsychological assessment. In conclusion, these results provide evidence for inter-areal CFS and PAC being two distinct mechanisms for coupling oscillations across frequencies in large-scale brain networks.

show abstract

Section: Large-scale Cfc Network Characterize Human Resting-state Brsupporting

confidence: 88%

Section: Introductionmentioning

confidence: 99%

Genuine cross-frequency coupling networks in human resting-state electrophysiological recordings

Siebenhuehner

Wang

Arnulfo

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

“…During task performance, PAC has been suggested to reflect the regulation of sensory information processing in β-and γ-frequencies by excitability fluctuations imposed by θ and α oscillations [9,10,34,36,37,42]. A large number of studies have identified local PAC, i.e., PAC observed between different frequency bands of the same signal, between the phases of slower oscillations in delta-(δ, 1-3 Hz), theta-(θ, 3-7 Hz), and α-frequency bands and the amplitude of γ oscillations in local field potentials (LFPs) in rats [43][44][45][46][47] and in human intracranial EEG [48][49][50][51] and MEG data [52][53][54][55][56]. Such local PAC has cortex-wide spatial modes akin to RSNs [53].…”

Section: Introductionmentioning

confidence: 99%

Genuine cross-frequency coupling networks in human resting-state electrophysiological recordings

et al. 2020

View full text Add to dashboard Cite

Phase synchronization of neuronal oscillations in specific frequency bands coordinates anatomically distributed neuronal processing and communication. Typically, oscillations and synchronization take place concurrently in many distinct frequencies, which serve separate computational roles in cognitive functions. While within-frequency phase synchronization has been studied extensively, less is known about the mechanisms that govern neuronal processing distributed across frequencies and brain regions. Such integration of processing between frequencies could be achieved via cross-frequency coupling (CFC), either by phase-amplitude coupling (PAC) or by n:m-cross-frequency phase synchrony (CFS). So far, studies have mostly focused on local CFC in individual brain regions, whereas the presence and functional organization of CFC between brain areas have remained largely unknown. We posit that interareal CFC may be essential for large-scale coordination of neuronal activity and investigate here whether genuine CFC networks are present in human resting-state (RS) brain activity. To assess the functional organization of CFC networks, we identified brain-wide CFC networks at mesoscale resolution from stereoelectroencephalography (SEEG) and at macroscale resolution from source-reconstructed magnetoencephalography (MEG) data. We developed a novel, to our knowledge, graphtheoretical method to distinguish genuine CFC from spurious CFC that may arise from nonsinusoidal signals ubiquitous in neuronal activity. We show that genuine interareal CFC is present in human RS activity in both SEEG and MEG data. Both CFS and PAC networks coupled theta and alpha oscillations with higher frequencies in large-scale networks connecting anterior and posterior brain regions. CFS and PAC networks had distinct spectral patterns and opposing distribution of low-and high-frequency network hubs, implying that they constitute distinct CFC mechanisms. The strength of CFS networks was also predictive of cognitive performance in a separate neuropsychological assessment. In conclusion,

show abstract

“…Apart from alpha activity, oscillatory activity in the theta (5-7 Hz) and gamma (60-80 Hz) frequency bands have also been linked to working memory. It has been proposed that theta-band oscillations underlie the organization of sequentially ordered working memory items, whereas gamma-band oscillations are thought to contribute to the maintenance of working memory information [24,[26][27][28][29][30].…”

Section: Introductionmentioning

confidence: 99%

Interpreting neural decoding models using grouped model reliance

2020

View full text Add to dashboard Cite

Machine learning algorithms are becoming increasingly popular for decoding psychological constructs based on neural data. However, as a step towards bridging the gap between theory-driven cognitive neuroscience and data-driven decoding approaches, there is a need for methods that allow to interpret trained decoding models. The present study demonstrates grouped model reliance as a model-agnostic permutation-based approach to this problem. Grouped model reliance indicates the extent to which a trained model relies on conceptually related groups of variables, such as frequency bands or regions of interest in electroencephalographic (EEG) data. As a case study to demonstrate the method, random forest and support vector machine models were trained on within-participant single-trial EEG data from a Sternberg working memory task. Participants were asked to memorize a sequence of digits (0-9), varying randomly in length between one, four and seven digits, where EEG recordings for working memory load estimation were taken from a 3-second retention interval. The present results confirm previous findings insofar as both random forest and support vector machine models relied on alpha-band activity in most subjects. However, as revealed by further analyses, patterns in frequency and topography varied considerably between individuals, pointing to more pronounced inter-individual differences than previously reported.

show abstract

Serial representation of items during working memory maintenance at letter-selective cortical sites

Cited by 104 publications

References 90 publications

Genuine cross-frequency coupling networks in human resting-state electrophysiological recordings

Genuine cross-frequency coupling networks in human resting-state electrophysiological recordings

Genuine cross-frequency coupling networks in human resting-state electrophysiological recordings

Interpreting neural decoding models using grouped model reliance

Contact Info

Product

Resources

About