Stimulus-dependent spiking relationships with the EEG

A. C. Snyder, Matthew A. SmithJournal of Neurophysiology 2015 · DOI

TL;DR

This study bridges the gap between non-invasive EEG and direct neuronal recordings by simultaneously measuring both in macaque visual cortex. They found that individual neurons' relationships with the EEG signal become more diverse and idiosyncratic during visual stimulation, suggesting these relationships reflect distinct computational roles within neural circuits.

What Problem Does This Paper Address?

Electroencephalography (EEG) is a cornerstone of non-invasive neuroscience, offering excellent temporal resolution for studying brain activity. However, its major limitation is poor spatial resolution, making it difficult to pinpoint the specific neural computations underlying the recorded signals. While we know EEG reflects the synchronized activity of large neuronal populations, understanding how individual neurons contribute to this macroscopic signal has been challenging due to a lack of studies simultaneously recording both single-neuron activity and EEG. This paper aims to address this by investigating the dynamic relationship between individual neuronal spiking and scalp EEG signals, particularly how this relationship changes between spontaneous and visually stimulated states.

Methods

The researchers employed a sophisticated experimental setup involving two adult male rhesus macaques. They implanted a 100-electrode "Utah" array in visual area V4 to record neuronal spiking activity and simultaneously recorded scalp EEG using eight Ag/AgCl electrodes.

Key analytical techniques included:

  • Spike-Triggered Averaging (STA): This method calculates the average EEG signal time-locked to individual neuronal spikes. It helps identify temporal patterns in the EEG that are associated with neuronal firing.
  • Spike Field Coherence (SFC): This technique quantifies the degree to which a neuron's spike train is phase-aligned with specific frequencies in the EEG signal. It measures the strength and preferred phase of this relationship.
  • Multiunit Activity (MUA) and Broadband EEG Correlation: They also examined the correlation between the aggregate spiking activity of neuronal populations (MUA) and the broadband EEG signal.
  • LFP Amplitude/Coherence and EEG Amplitude Correlation: The study further investigated how local field potential (LFP) amplitude and coherence relate to EEG amplitude across different frequency bands.
  • Trial-Shuffling Correction: To mitigate potential confounds arising from collecting spontaneous and evoked data in separate sessions, a trial-shuffling procedure was used to estimate and subtract measurement biases.
  • Phase Consistency Analysis: This method assessed how consistently individual neurons aligned to the preferred EEG phase, controlling for coherence strength to avoid spurious findings.

Datasets

The study utilized data collected from two adult male rhesus macaques. For the spontaneous condition, data were collected over 14 sessions for monkey B and 10 sessions for monkey W. For the visually evoked condition, data were collected over 10 sessions for monkey B and 8 sessions for monkey W. The paper does not introduce new benchmark datasets.

Key Insights and Ideas

  • EEG as a Window into Population Dynamics: The study reinforces the idea that EEG primarily reflects the synchronized activity of large neuronal populations, but it also highlights that the nature of this synchronization changes with cognitive state.
  • Idiosyncratic Neuron-EEG Relationships: A central finding is that individual neurons don't just passively contribute to the EEG; they have distinct, often idiosyncratic, relationships with it. These relationships are not uniform across all neurons.
  • Stimulus-Dependent Shift in Coordination: Visual stimulation fundamentally alters how neurons relate to the EEG. They shift from a state where they are more uniformly embedded in global network activity (spontaneous) to a state where their unique computational roles become more prominent.
  • Computational Roles Reflected in EEG Coupling: The diversity of spike-EEG relationships during stimulation suggests that these patterns may serve as a signature of a neuron's specific computational function within its local circuit.
  • Low-Frequency Dominance: The strongest relationships between spiking activity and EEG were predominantly observed in lower frequency bands (delta, theta, alpha), suggesting these oscillations are key carriers of information about population-level activity.

Key Results

  • Spike-Triggered Average (STA): In the spontaneous state, the grand-averaged STA showed a clear negative deflection before a spike and a positive deflection after. This pattern was present but less pronounced and delayed during visual stimulation.
  • Spike Field Coherence (SFC): SFC was significantly reduced in the stimulus-evoked condition compared to the spontaneous condition, particularly below 30 Hz. This indicates a weaker phase-locking between individual neuron spikes and the EEG signal when stimulated.
  • Phase Consistency: When controlling for coherence strength, phase consistency of spike-EEG relationships was significantly reduced during visual stimulation, especially in the 7-11 Hz range. This means neurons aligned to more diverse EEG phases when stimulated.
  • LFP-EEG Correlation: LFP amplitude was significantly correlated with EEG amplitude across most frequencies in the spontaneous state. During stimulation, LFP coherence (but not amplitude) became correlated with EEG amplitude at low frequencies.
  • Significant STAs: A substantial proportion of neurons (67.1%) exhibited significantly nonzero STAs, indicating a measurable relationship between their spiking and the EEG.

Limitations and Future Work

  • Craniotomy: The study was conducted on macaques with craniotomies, which could potentially alter the volume conductor properties of the brain. However, the authors cite previous work suggesting negligible effects of craniotomies on EEG signals, lending confidence to the generalizability of their findings.
  • Gamma Band: The study did not find significant gamma-band SFC or induced amplitude increases with visual stimulation, contrary to some previous reports. The authors suggest this might be due to specific experimental parameters, the recording area (V4 vs. V1), or the depth of their electrodes. They propose that gamma-band effects might be more context-specific than previously assumed.
  • Future Work: The authors suggest that understanding the specific ways individual neurons relate to EEG signals may hold clues about their computational roles. Future research could explore how these diverse spike-EEG relationships map onto known functional specializations within visual cortex. They also highlight the potential of concurrent EEG and intracranial recordings to provide substantial information about global context for local population activity.

Summary updated

Abstract

The development and refinement of noninvasive techniques for imaging neural activity is of paramount importance for human neuroscience. Currently, the most accessible and popular technique is electroencephalography (EEG). However, nearly all of what we know about the neural events that underlie EEG signals is based on inference, because of the dearth of studies that have simultaneously paired EEG recordings with direct recordings of single neurons. From the perspective of electrophysiologists there is growing interest in understanding how spiking activity coordinates with large-scale cortical networks. Evidence from recordings at both scales highlights that sensory neurons operate in very distinct states during spontaneous and visually evoked activity, which appear to form extremes in a continuum of coordination in neural networks. We hypothesized that individual neurons have idiosyncratic relationships to large-scale network activity indexed by EEG signals, owing to the neurons' distinct computational roles within the local circuitry. We tested this by recording neuronal populations in visual area V4 of rhesus macaques while we simultaneously recorded EEG. We found substantial heterogeneity in the timing and strength of spike-EEG relationships and that these relationships became more diverse during visual stimulation compared with the spontaneous state. The visual stimulus apparently shifts V4 neurons from a state in which they are relatively uniformly embedded in large-scale network activity to a state in which their distinct roles within the local population are more prominent, suggesting that the specific way in which individual neurons relate to EEG signals may hold clues regarding their computational roles.
BibTeX
@article{snyder2015stimulus,
  author = {A. C. Snyder and Matthew A. Smith},
  title = {{Stimulus-dependent spiking relationships with the EEG}},
  journal = {Journal of Neurophysiology},
  year = {2015},
  doi = {10.1152/jn.00427.2015},
}
Figure