Topological structure of population activity in mouse visual cortex encodes densely sampled stimulus rotations

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• The primary visual cortex (V1) is crucial for visual processing and sensory computation.
• High-density neural recordings reveal that neural population activity often lies on low-dimensional manifolds.

Purpose of the Study:

• To quantify the geometric structure of neural manifolds in the primary visual cortex.
• To investigate how neural population activity encodes visual stimuli, specifically rotation angle.

Main Methods:

• Analysis of publicly available two-photon optical recordings from mouse V1.
• Utilizing a geodesic metric and persistent homology to analyze neural population activity.
• Examining responses to visual stimuli with a densely sampled rotation angle.

Main Results:

• Neural population activity in V1 forms a circular manifold in response to rotating visual stimuli.
• This circular manifold effectively encodes the angle of visual rotation.
• The expression of this circular manifold differs among neuronal subpopulations with distinct orientation and direction selectivity.

Conclusions:

• The structure of neural population activity in V1 exhibits a discernible topology, specifically a circular manifold.
• Understanding this manifold structure provides insights into how the brain represents sensory information.
• Challenges exist in accurately reconstructing the true topology of neural population dynamics.