Hierarchical Neural Circuit Theory of Normalization and Inter-areal Communication

Area of Science:

• Computational Neuroscience
• Systems Neuroscience
• Neural Circuit Theory

Background:

• The primate brain's hierarchical, modular architecture features conserved microcircuits.
• Feedback connections are ubiquitous but their precise functions remain largely unknown.
• Understanding feedback is crucial for deciphering inter-areal communication and brain function.

Purpose of the Study:

• To investigate the functional role of feedback connections in hierarchical neural circuits.
• To develop a theoretical framework explaining divisive normalization and inter-areal communication.
• To generate experimentally-testable predictions regarding neural dynamics and connectivity.

Main Methods:

• Development of a hierarchical neural circuit theory incorporating feedback connections.
• Analytical derivation of power and coherence spectra for a two-stage V1-V2 model.
• Validation of theoretical predictions against experimental observations and multi-area simulations.

Main Results:

• The theory dynamically implements divisive normalization, amplifying responses consistent with experimental data.
• Derived spectra match experimental observations, showing frequency shifts with stimulus contrast and 1/f^4 decay.
• Feedback enhances inter-areal communication and diminishes within-area communication, predicting communication subspaces and coherence bands.

Conclusions:

• The proposed theory offers a unified, analytically tractable framework for understanding normalization and brain connectivity.
• Feedback strength modulates inter-areal communication and functional connectivity, impacting subspace dimensionality and coherence.
• The theory provides novel, testable predictions for future experimental investigations into neural dynamics.