Recurrent competition explains temporal effects of attention in MSTd

Area of Science:

• Neuroscience
• Computational Neuroscience
• Visual Perception

Background:

• Radial optic flow fields, characterized by a focus of expansion (FoE), indicate an observer's heading direction.
• Cells in the dorsal medial superior temporal area (MSTd) of primates are sensitive to these fields and are believed to be heading-sensitive.
• Human navigation involves dynamic shifts in attention, influenced by environmental context and object interactions.

Purpose of the Study:

• To develop a dynamical neural circuit model that replicates electrophysiologically observed phenomena in primate MSTd neurons related to heading perception and attention.
• To investigate how attentional signals modulate the spatial tuning and population activity of heading-sensitive neurons.
• To elucidate the computational mechanisms underlying heading perception under dynamic attentional conditions.

Main Methods:

• A dynamical neural circuit model was constructed, incorporating components analogous to MT (motion pooling) and MSTd (recurrent competition, attention integration from FEF).
• The model simulates radial optic flow fields and integrates attentional signals with varying spatial relationships to the FoE.
• Model outputs, including neuron tuning curves and population activation profiles, were compared against neurophysiological data.

Main Results:

• The model successfully reproduced the linear temporal shift in peak population activity observed in primate MSTd neurons as the attentional prime-FoE distance increased.
• The model's neuron tuning curves and population activation profiles qualitatively matched experimental findings.
• The study found that the precise nature of attention (multiplicative vs. non-multiplicative) was less critical than its Gaussian-like profile for model performance.

Conclusions:

• The developed neural circuit model provides a plausible mechanism for how MSTd neurons process heading information influenced by attention.
• The model's ability to replicate key experimental findings supports its validity in explaining heading perception dynamics.
• Further predictions suggest that observed population activity deflections may indicate shifts in attentional balance between priming and FoE locations.