The geometry and dimensionality of brain-wide activity

Area of Science:

• Neuroscience
• Computational Biology
• Systems Neuroscience

Background:

• Understanding how neural activity is organized across the brain is crucial for deciphering brain function.
• Previous studies have explored neural activity patterns, but a unifying principle for brain-wide organization remained elusive.

Purpose of the Study:

• To investigate the organizational principles of whole-brain neural activity.
• To determine if neural activity organization exhibits scale-invariance across different scales of neuronal populations.

Main Methods:

• Recorded whole-brain calcium activity in larval zebrafish during distinct behavioral states (hunting and spontaneous).
• Analyzed the neural covariance spectrum to describe the shape of the neural activity space.
• Applied Euclidean Random Matrix theory to model the observed neural activity patterns and scale-invariance.

Main Results:

• The neural activity space exhibits scale-invariance: smaller, randomly sampled cell assemblies statistically resemble the entire brain's activity space.
• This scale-invariance is explained by neurons reorganizing from anatomical to functional positions based on correlations.
• Key contributing factors include slow neural correlation decay, a high functional space dimension, and neural activity heterogeneity.

Conclusions:

• A scale-invariant organization principle governs brain-wide neural activity, applicable across different species like zebrafish and mice.
• The geometry of neural activity space evolves predictably with population size and sampling methods.
• This study reveals a fundamental organizing principle of brain-wide neural activity, linking neural correlations to functional organization.