Directed information exchange between cortical layers in macaque V1 and V4 and its modulation by selective attention

Demetrio Ferro, Jochem van Kempen, Michael Boyd, Stefano Panzeri, Alexander Thiele

TL;DR

This study investigated how attention shapes information flow between and within cortical layers of macaque V1 and V4 using advanced signal analysis. Surprisingly, attention reduced communication within V1 while increasing it between V1 and V4, challenging existing theories of frequency-specific neural communication and highlighting the complex interplay of top-down and bottom-up processing.

What Problem Does This Paper Address?

The brain needs to integrate sensory input with cognitive goals to guide behavior. This integration is thought to involve complex communication networks within and between different cortical areas and even within the distinct layers of the cortex. While feedforward connections are generally associated with sensory input and feedback with cognitive control, and specific frequency bands (like gamma for feedforward and alpha/beta for feedback) have been implicated, the precise mechanisms and layer-specific dynamics of this information exchange, especially under attentional modulation, remain poorly understood. This paper aims to map this flow of information, specifically focusing on how selective attention alters communication patterns between cortical layers in visual areas V1 and V4.

Methods

The researchers employed local field potential (LFP) recordings from macaque monkeys performing a visual attention task. To quantify the direction and strength of information flow, they utilized spectrally resolved conditional Granger causality (cGC). This advanced technique allows for the analysis of directed causal influences between time series (in this case, LFPs from different cortical depths and areas) across various frequency bands. They also analyzed spectral power and coherence to understand how attention affects the overall activity and synchronization within and between areas. To account for potential confounds, they used bipolar referencing for LFPs and rigorous statistical methods, including FDR correction.

Datasets

The study used electrophysiological data recorded from two adult male rhesus macaque monkeys performing a feature-based spatial attention task. The data comprised simultaneous LFP recordings from visual areas V1 and V4 across multiple cortical layers. No new datasets were introduced; the study relies on novel analysis of existing experimental data.

Key Insights and Ideas

  • Attention is a double-edged sword for local communication: While attention is often thought to enhance neural processing, it surprisingly decreased communication within V1 layers, except for a specific granular-to-supragranular pathway. This suggests attention might be "pruning" or simplifying local processing to focus resources elsewhere.
  • Inter-area communication gets a boost: In contrast to V1, attention significantly increased communication within V4 and, crucially, between V1 and V4, particularly in the feedforward direction. This highlights how attention orchestrates communication between brain regions.
  • Frequency specificity is more complex than thought: Existing theories often link specific frequency bands to feedforward (gamma) and feedback (alpha/beta) pathways. This study found that while gamma was indeed stronger in the V1-to-V4 feedforward direction, attention modulated V4-to-V1 feedback across all frequencies, and increased V1-to-V4 communication across all frequency bands, challenging strict frequency-band segregation.
  • Layer-specific dynamics are nuanced: Information flow within V1 predominantly moved upwards towards output layers, while V4 showed more complex bidirectional flow. Attention's impact also varied by layer, but not always in ways predicted by simple feedforward/feedback models.

Key Results

  • Within V1: Attention decreased cGC between most layer pairs, except for granular-to-supragranular interactions.
  • Within V4: Attention generally increased cGC across all frequencies.
  • V1 to V4 (Feedforward): Attention increased cGC across all frequency bands.
  • V4 to V1 (Feedback): Attention decreased low-frequency feedback (theta/alpha) but increased beta and gamma feedback.
  • Dominant flow: Before attention, V1 showed upward flow, V4 had bidirectional flow, V1-to-V4 was stronger in gamma, and V4-to-V1 was stronger in lower frequencies.

Limitations and Future Work

The authors acknowledge that their findings challenge existing theories, particularly regarding the strict frequency-band segregation of feedforward and feedback signals. They also note that the indirect pathways (e.g., via area V2) could play a role in inter-areal communication, which their direct V1-V4 analysis doesn't fully capture. Future work could explore these indirect routes and further investigate the precise mechanisms by which attention modulates these layer-specific and frequency-dependent interactions, potentially by combining cGC with other techniques like optogenetics or single-unit recordings. The study also raises questions about the role of inhibitory processes and predictive coding in these attentional effects, suggesting avenues for further theoretical and experimental exploration.

Summary updated

Abstract

Achieving behavioral goals requires integration of sensory and cognitive information across cortical laminae and cortical regions. How this computation is performed remains unknown. Using local field potential recordings and spectrally resolved conditional Granger causality (cGC) analysis, we mapped visual information flow, and its attentional modulation, between cortical layers within and between macaque brain areas V1 and V4. Stimulus-induced interlaminar information flow within V1 dominated upwardly, channeling information toward supragranular corticocortical output layers. Within V4, information flow dominated from granular to supragranular layers, but interactions between supragranular and infragranular layers dominated downwardly. Low-frequency across-area communication was stronger from V4 to V1, with little layer specificity. Gamma-band communication was stronger in the feedforward V1-to-V4 direction. Attention to the receptive field of V1 decreased communication between all V1 layers, except for granular-to-supragranular layer interactions. Communication within V4, and from V1 to V4, increased with attention across all frequencies. While communication from V4 to V1 was stronger in lower-frequency bands (4 to 25 Hz), attention modulated cGCs from V4 to V1 across all investigated frequencies. Our data show that top-down cognitive processes result in reduced communication within cortical areas, increased feedforward communication across all frequency bands, and increased gamma-band feedback communication.
BibTeX
@inproceedings{ferro2021directed,
  author = {Demetrio Ferro and Jochem van Kempen and Michael Boyd and Stefano Panzeri and Alexander Thiele},
  title = {{Directed information exchange between cortical layers in macaque V1 and V4 and its modulation by selective attention}},
  booktitle = {Proceedings of the National Academy of Sciences},
  year = {2021},
  doi = {10.1073/pnas.2022097118},
}
Figure