Orientation Selectivity from Very Sparse LGN Inputs in a Comprehensive Model of Macaque V1 Cortex

Area of Science:

• Computational neuroscience
• Systems neuroscience
• Primate visual system modeling

Background:

• The primary visual cortex (V1) in macaque monkeys receives sparse input from the Lateral Geniculate Nucleus (LGN).
• Explaining V1 functions like orientation selectivity with sparse input has been a theoretical challenge.
• Previous models often overlooked the significant role of intracortical interactions.

Purpose of the Study:

• To develop a computational model of macaque V1 that reconciles sparse LGN input with observed visual functions.
• To investigate how robust orientation selectivity and map continuity can emerge from sparse inputs.
• To explore dynamic regimes of the model that emulate diverse V1 phenomena and experimental data.

Main Methods:

• Construction of a novel computational model of the macaque primary visual cortex (V1).
• Incorporation of anatomically realistic, sparse magnocellular LGN input data.
• Analysis of model dynamics to emulate orientation selectivity, neuronal response diversity, and gamma-band oscillations.

Main Results:

• The model successfully generates robust orientation selectivity and orientation map continuity despite sparse LGN input.
• Dynamic regimes were identified that simultaneously emulate multiple V1 phenomena, including neuronal response diversity and modulation ratios.
• Intracortical interactions were shown to be fundamental to all modeled visual functions of V1.

Conclusions:

• Sparse LGN input is sufficient for V1 orientation selectivity, with intracortical interactions playing a dominant role.
• Cortical population dynamics, driven by intracortical signals, are essential for V1 function and cannot be explained by single neurons alone.
• The study advocates for a paradigm shift in neuroscience modeling towards comprehensive, data-driven models emphasizing population dynamics.