Visual areas exert feedforward and feedback influences through distinct frequency channels

By André Moraes Bastos, Julien Vezoli, Conrado A. Bosman, Jan-Mathijs Schoffelen, Robert Oostenveld, J. R. Dowdall, P. Weerd, Henry Kennedy, Pascal Fries. bioRxiv, 2014

1,270 citations recorded by the source corpus.

Visual cortical areas are thought to form a hierarchy and to subserve cognitive functions by interacting in both feedforward and feedback directions1. While feedforward influences convey sensory signals, feedback influences modulate brain responses to a given sensory stimulus according to the current behavioural context. Many studies have demonstrated effects of feedback influences on feedforward driven responses2 and on behaviour3. Also, anatomical projections in both directions have been identified1, 4. However, although these studies have revealed the anatomical paths and the neurophysiological consequences of influences in both directions, the neurophysiological mechanisms through which these influences are exerted remain largely elusive. Here we show that in the primate visual system, feedforward influences are carried by theta-band (∼4 Hz) and gamma-band (∼60-80 Hz) synchronization, and feedback influences by beta-band (∼14-18 Hz) synchronization. These frequency-specific asymmetries in directed influences were revealed by simultaneous local field potential recordings from eight visual areas and an analysis of Granger-causal influences across all 28 pairs of areas. The asymmetries in directed influences correlated directly with asymmetries in anatomy and enabled us to build a visual cortical hierarchy from the influence asymmetries alone. Across different task periods, most areas stayed at their hierarchical position, whereas particularly frontal areas moved dynamically. Our results demonstrate that feedforward and feedback signalling use different frequency channels, which might subserve their differential communication requirements and lead to differential local consequences. The possibility to infer hierarchical relationships through functional data alone might make it possible to derive a cortical hierarchy in the living human brain.

Subfields: Cellular and Molecular Neuroscience, Cognitive Neuroscience

View Paper: https://www.semanticscholar.org/paper/9d6d578dfe1753b8f3e9514bcb1f4037c88fab76

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cellular and Molecular Neuroscience

In Cognitive Neuroscience

Citations in the corpus

Alpha-beta and gamma rhythms subserve feedback and feedforward influences among human visual cortical areas (2016, positive): “Given that the macaque SLN is correlated to the human DAI, we asked whether the human DAI also establishes a hierarchy of human visual areas, similar to the DAI-based hierarchy of macaque visual areas established previously (Bastos et al., 2015).”
The graphical brain: Belief propagation and active inference (2017, background): “…and the rates of change of expectations are functions of prediction errors, one would expect to see the same sort of spectral differences in laminar-speciﬁc oscillations that have been previously discussed in the context of predictive coding (Bastos et al., 2012, 2015; Bosman et al., 2012).”
Prefrontal Cortex Networks Shift from External to Internal Modes during Learning (2016, background): “In contrast, we observed little evidence for a role for gamma oscillations in our data, in line with previous work suggesting gamma oscillations are less prominent in the frontal cortex (Siegel et al., 2009; Antzoulatos and Miller, 2014) than in the posterior visual cortex (Bastos et al., 2015).”
Switching between internal and external modes: A multiscale learning principle (2017, background): “a more prior-driven, top-down (internal) mode of processing (Bastos et al., 2015; Halgren et al., 2015; Sherman, Kanai, Seth, & VanRullen, 2016; Van Kerkoerle et al., 2014).”
Toward an Integration of Deep Learning and Neuroscience (2016, background): “Feedforwardandfeedbackpathwaysarealsodistinguisheddynamically,e.g.,bypropagating diﬀerentoscillatoryfrequencies(VanKerkoerleetal.,2014;Bastosetal.,2015),andmoleculary,e.g.,withNMDAreceptorsplayinganimportantroleinfeedback processing. such sub-networks might even be part of the same neuron and map…”
Brain Mechanisms of Concept Learning (2019, background): “…(bottom-up vs top-down) for category abstraction in the PFC. Gamma oscillations are typically linked to the feedforward flow of cortical activation (Bastos et al., 2015; Jensen et al., 2015) and the vlPFC receives direct inputs from inferior temporal cortex (O’Reilly, 2010), potentially continuing…”
Rhythms For Cognition: Communication Through Coherence (2015, positive): “…even though the areas are bidirectionally coupled, this gamma-band coherence does not occur at zero phase, but with a systematic delay, i.e. with a directedness (Bastos et al., 2015b; Bosman et al., 2012; Gregoriou et al., 2009; Grothe et al., 2012; Jia et al., 2013a; Zandvakili and Kohn, 2015).”
Granger Causality Analysis in Neuroscience and Neuroimaging (2015, background): “For example, recent applications to monkey electrophysiological data have revealed “top-down” G-causality influences in the alpha and beta ranges, and “bottom-up” influences in the gamma range, consistent with popular predictive processing frameworks (Bressler and Richter, 2014; van Kerkoerle et al.”
Alpha and gamma oscillations characterize feedback and feedforward processing in monkey visual cortex (2014, result): “The feedforward signature of γ is in line with recent studies showing stronger Granger causality for γ from V1 to V2 and from V1 to V4 than in the opposite direction (18, 49, 65).”
Conscious Processing and the Global Neuronal Workspace Hypothesis (2020, background): “, 2016) leads to the emergence of frequency-bandspecific patterns of causality (bottom-up gamma arising primarily in supragranular layers versus top-down alpha-beta arising primarily from bottom layers), which have been observed empirically across the cortical hierarchy in both human and non-human p”
Neural mechanisms of time-forward predictions for naturalistic auditory tone sequences (2026, unknown): “Early electrophysiological studies supporting the predictive coding framework found that predictions were conveyed by low-frequency signals in the top-down direction and PEs were conveyed by high-frequency signals in the bottom-up direction 5 – 7, 36,37 .”
Distinct laminar origins of sensory-evoked high-gamma and low-frequency ECoG signals revealed by optogenetics (2026, unknown): “This is distinct from oscillations, which may emerge from feedback interactions and may contribute to interareal communication, and which are often inferred from spectral structure in the ECoG and LFP literature (2, 64, 65).”