Analytic Model for Feature Maps in the Primary Visual Cortex.

Area of Science:

• Computational Neuroscience
• Neuroscience
• Visual System Modeling

Background:

• The primary visual cortex (V1) exhibits complex spatial maps of orientation preference (OP) and ocular dominance (OD).
• Understanding the interplay between OP and OD and their spatial organization is crucial for comprehending visual processing.
• Existing models often lack the compactness and analytical tractability needed for large-scale neural simulations.

Purpose of the Study:

• To propose a compact analytic model for the combined OP-OD map in V1.
• To investigate the mutual constraints between OP and OD features and their impact on spatial layout.
• To provide a framework for analyzing V1's OP-OD structure using neural field theory.

Main Methods:

• Development of an anisotropic Laplacian (AL) operator for neural sensitivity to visual input orientation.
• Integration of a receptive field (RF) operator modeling anisotropic spatial projections from nearby neurons.
• Construction of a periodic map describing operator parameter variations across V1, fitted to experimental OP tuning curves.

Main Results:

• RF anisotropy, not AL anisotropy, primarily determines OP selectivity, aligning with Hubel and Wiesel's findings on RF elongation and tuning width.
• A simplified OP-OD map model demonstrates that dominant spatial Fourier coefficients suffice to reconstruct the basic map structure.
• Analysis of simulated V1 structures reveals that map irregularities correlate with the spread of dominant Fourier coefficients.

Conclusions:

• The proposed compact analytic model effectively captures the combined OP-OD structure of V1.
• The model's reliance on RF anisotropy provides a clear explanation for orientation tuning properties.
• The Fourier-based representation facilitates compact analysis and integration into neural field theory for V1 activity modulation.