Gain Modulation as a Mechanism for Coding Depth from Motion Parallax in Macaque Area MT

Area of Science:

• Neuroscience
• Computational Vision
• Sensory Processing

Background:

• Motion parallax (MP) is a monocular depth cue from observer translation.
• MP ambiguity in depth sign (near vs. far) requires integrating retinal motion with eye movement signals.
• Previous work showed extra-retinal and visual signals from smooth eye movements modulate MT neuron responses, creating depth-sign selectivity.

Purpose of the Study:

• To elucidate the neural mechanisms underlying depth-sign selectivity in area MT.
• To analyze MT neuron responses based on retinal velocity and eye movement direction.
• To determine how smooth eye movements generate depth-sign selectivity.

Main Methods:

• Analysis of macaque MT neuron responses to retinal velocity and eye movement direction.
• Investigating the modulatory effects of smooth eye movements on MT neuron responses.
• Computational simulations to assess depth estimation from MT population activity.

Main Results:

• Smooth eye movements systematically modulate MT responses in a temporally precise and directionally specific manner.
• Depth-sign selectivity is primarily achieved through multiplicative modulation of MT neuron response gain.
• Simulations indicate reasonable depth estimation via linear decoding of MT population with eye-velocity-dependent gains.

Conclusions:

• This study provides the first mechanistic description of how visual cortical neurons signal depth from motion parallax.
• Smooth eye movements modulate MT neuron response gain directionally, generating depth-sign selectivity.
• Depth can be estimated from a population of gain-modulated MT neurons.