Measuring Stimulus Information Transfer Between Neural Populations through the Communication Subspace.

Area of Science:

• Computational Neuroscience
• Systems Neuroscience
• Information Theory

Background:

• Sensory information processing relies on neural communication across brain areas.
• Shared neural response variability within populations limits stimulus information representation.
• The effect of this variability on interareal communication remains unclear.

Purpose of the Study:

• To develop a mathematical framework for understanding neural population response variability's impact on sensory information transmission.
• To investigate how variability affects interareal communication in the brain.

Main Methods:

• Combined linear Fisher information with the communication subspace framework.
• Partitioned Fisher information based on the alignment of population covariance and mean tuning direction.
• Performed mathematical and numerical analyses of the proposed Fisher information decomposition.

Main Results:

• Developed a method to partition Fisher information, revealing how variability influences information transmission.
• Demonstrated theoretical scenarios for flexible routing and gating of sensory information using communication subspaces.
• Quantified the relationship between neural variability and the fidelity of interareal communication.

Conclusions:

• The proposed framework provides a theoretical lens for understanding sensory information transmission between brain areas.
• This work guides experimental design for investigating interareal communication.
• Highlights the role of communication subspaces in modulating information flow modulated by neural variability.