Neural Noise

Brain-wide cellular-resolution map of neural activity (e.g. decision-making in mice)

2 min read

  • An international collaboration including MIT’s Ila Fiete produced a brain-wide map of decision-making at cellular resolution in mice.1
  • The map supports neural activity mapping, decoding, and computational neuroscience applications.1 1

Gardner updates

  • A brain-wide map of neural activity during complex behaviour was reported in Nature, supporting neural data analysis and decoding and foundational for understanding dynamics. 2

  • An international collaboration including MIT built a brain-wide map of decision-making at cellular resolution in mice. 1

  • A Nature report described a brain-wide map of neural activity during complex behaviour, supporting neural data analysis and decoding; likely animal model. 2

Weekly enrichment (2026-07-20)

  • The map is the work of the International Brain Laboratory (IBL) — a consortium of labs across Europe and the U.S. including MIT’s Ila Fiete — and was published on 3 September 2025 as two open-access papers in Nature.34
  • The dataset comprises 621,733 neurons recorded with 699 Neuropixels probe insertions across 139 mice in 12 laboratories, covering 279 brain areas that together represent about 95% of the mouse brain volume.53
  • After spike sorting (Kilosort with custom additions), the recordings yielded 75,708 well-isolated neurons, averaging about 108 per probe.5
  • The behavioral task had sensory, motor, and cognitive components: a mouse turned a wheel toward a visual stimulus appearing on the left or right for reward, with faint stimuli forcing use of a block-structured prior (probability alternating around 0.2 and 0.8).35
  • The first paper, “A brain-wide map of neural activity during complex behaviour,” found decision-making signals are broadly distributed rather than localized: stimulus representations spread from classical visual areas into midbrain/hindbrain regions, motor-action-correlated responses appeared almost everywhere, and reward responses were widespread.5
  • The second paper, “Brain-wide representations of prior information,” reported that prior expectations are encoded across the brain, including early sensory relays such as the thalamus, supporting a view of the brain as a prediction machine and with possible relevance to schizophrenia and autism.3
  • The findings challenge the traditional strictly hierarchical model of brain information processing, emphasizing constant cross-region communication during decision-making, movement onset, and reward.3
  • IBL used a standardized, shared toolset and data-processing pipelines across labs (a CERN/Human Genome Project-style model), and all data, tools, and protocols are openly released for community reuse.34

Footnotes

  1. https://news.mit.edu/2025/comprehensive-cellular-resolution-map-brain-activity-0904 ↩ ↩2 ↩3 ↩4

  2. https://news.google.com/rss/articles/CBMiX0FVX3lxTE5nUW9WVm9zWC13NEhpNnByOFNXQmVOdllDMU8zWWo4cE9SdEwtLWdrcTlmc0d6cENjMmo0NWUzRG54aVBnZXlOY0Q4TVZEc2FsYWJKZ1E5djhVTVpKb1Bn?oc=5 ↩ ↩2

  3. https://www.internationalbrainlab.com/the-brainwide-map-press ↩ ↩2 ↩3 ↩4 ↩5 ↩6

  4. https://pni.princeton.edu/news/2025/first-brain-wide-map-decision-making-charted-mice ↩ ↩2

  5. https://www.nature.com/articles/s41586-025-09235-0 ↩ ↩2 ↩3 ↩4

Graph View