1024-channel electrophysiological recordings in macaque V1 and V4 during resting state

Xing Chen, Aitor Morales-Gregorio, Julia Sprenger, Alexander Kleinjohann, Shashwat Sridhar, Sacha J. van Albada, Sonja Grün, Pieter R. RoelfsemaScientific 2022 · DOI

TL;DR

Researchers have released an unprecedented dataset featuring simultaneous 1024-channel electrophysiological recordings from the visual cortex (V1 and V4) of macaques during resting states. This open-access resource bridges the gap between the broad-but-slow view of fMRI and the sharp-but-narrow view of traditional electrode recordings, offering a high-resolution map of "background" brain activity.

What Problem Does This Paper Address?

Even when you aren't looking at anything specific, your visual cortex is buzzing with activity. These "resting state correlations" are thought to reflect the brain's underlying functional architecture, feedback loops, and neuromodulatory shifts.

Until now, researchers studied this in two ways, both with major trade-offs:

  1. fMRI: Provides a whole-brain view but is "blurry" in time (seconds) and space (millimeters).
  2. Standard Electrophysiology: Provides incredible temporal detail (milliseconds) but usually only from a handful of electrodes, like looking at a stadium through a straw.

The challenge is technical and surgical: implanting over a thousand electrodes into a primate brain and maintaining signal quality over time is incredibly difficult. This paper describes a successful system to overcome these hurdles, providing a "high-definition" view of cortical dynamics.

Methods

The team developed a novel neuronal recording system and a customized, 3D-printed implantable interface.

  • The Hardware: They used 16 "Utah" electrode arrays (8x8 grids) for a total of 1024 channels. These were connected via 7-cm wire bundles to a custom pedestal on the skull.
  • The Surgery: The arrays were chronically implanted in the visual cortex (V1 and V4) of two rhesus macaques.
  • Signal Processing: They recorded raw data at 30 kHz. From this, they extracted two key signals:
  • LFP (Local Field Potentials): Low-frequency oscillations (1–100 Hz) reflecting integrated synaptic input.
  • MUAe (Multi-unit Activity envelope): High-frequency signals (500–9000 Hz) representing the aggregate spiking activity of neurons near the electrode tip.
  • Validation: They used infrared eye-tracking to correlate brain activity with whether the monkeys' eyes were open or closed—a critical factor in resting-state research.

Datasets

The authors have made the data available on the G-Node Infrastructure (GIN), following FAIR (Findable, Accessible, Interoperable, and Reusable) principles. The release includes:

  1. Resting State Data: The core dataset of spontaneous activity.
  2. SNR Task Data: Recordings during a checkerboard stimulus task used to calculate the Signal-to-Noise Ratio (SNR) for each channel.
  3. RF Mapping Data: Recordings during a moving bar task used to map the Receptive Fields (RFs) of the neurons, showing exactly which part of the "visual world" each electrode monitors.

Key Insights and Ideas

  • Spatiotemporal Resolution: This is the first dataset of its scale in non-human primates. It allows researchers to observe "brain waves" or traveling patterns of activity across large swaths of the cortex in real-time.
  • The "Synchrofact" Problem: A major technical insight involves "synchrofacts"—spurious correlations caused by electronic cross-talk or external electromagnetic noise. The authors developed a systematic way to detect and remove these artifacts using a "synchrofact participation" (SP) metric, ensuring the correlations studied are biological, not electrical noise.
  • Eye State Matters: The study confirmed that V1 activity levels are significantly higher when the eyes are open, even in a dark room, compared to when they are closed. This validates the use of their eye-tracking data as a vital metadata layer.

Key Results

  • Successful Scale: The system successfully recorded from 1024 channels simultaneously across two brain areas (V1 and V4).
  • High Signal Quality: In early sessions, nearly 90% of channels (899/1024) showed high-quality, stimulus-evoked responses (SNR ≥ 2).
  • Retinotopic Mapping: The RF mapping confirmed a clear spatial organization that matched the physical location of the arrays on the cortex, proving the electrodes were accurately sampling the visual field.
  • Longevity: The implants remained mechanically stable and provided usable data for years (up to 4 years for head posts and 1 year for the electrode pedestal).

Limitations and Future Work

  • Signal Decay: As with most chronic implants, the number of "good" channels with high SNR decreased over months and years due to tissue scarring (gliosis) or electrode degradation.
  • Cross-talk: While the authors provide a method to remove "synchrofacts," some electronic noise is inevitable in such high-density systems, requiring users of the data to be rigorous in their filtering.
  • Sample Size: The data comes from only two monkeys. While typical for primate research, individual variations in cortical folding and connectivity exist.
  • Future Directions: The authors suggest this dataset could be used to build large-scale computational models of the visual system or to develop new algorithms for detecting traveling waves in neural populations.

Summary updated

Abstract

Co-variations in resting state activity are thought to arise from a variety of correlated inputs to neurons, such as bottom-up activity from lower areas, feedback from higher areas, recurrent processing in local circuits, and fluctuations in neuromodulatory systems. Most studies have examined resting state activity throughout the brain using MRI scans, or observed local co-variations in activity by recording from a small number of electrodes. We carried out electrophysiological recordings from over a thousand chronically implanted electrodes in the visual cortex of non-human primates, yielding a resting state dataset with unprecedentedly high channel counts and spatiotemporal resolution. Such signals could be used to observe brain waves across larger regions of cortex, offering a temporally detailed picture of brain activity. In this paper, we provide the dataset, describe the raw and processed data formats and data acquisition methods, and indicate how the data can be used to yield new insights into the 'background' activity that influences the processing of visual information in our brain.
BibTeX
@inproceedings{chen2022channel,
  author = {Xing Chen and Aitor Morales-Gregorio and Julia Sprenger and Alexander Kleinjohann and Shashwat Sridhar and Sacha J. van Albada and Sonja Grün and Pieter R. Roelfsema},
  title = {{1024-channel electrophysiological recordings in macaque V1 and V4 during resting state}},
  booktitle = {Scientific},
  year = {2022},
  doi = {10.1038/s41597-022-01180-1},
}
Figure