Visual-Based Spatial Coordinate Dominates Probabilistic Multisensory Inference in Macaque MST-d Disparity Encoding.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Sensory Processing

Background:

• Multisensory integration and separation are critical for accurate environmental perception.
• The dorsal medial superior temporal area (MST-d) is implicated in processing multisensory information for motion perception.
• The reference frame used for multisensory integration remains an open question.

Purpose of the Study:

• To investigate whether MST-d neurons use visual or vestibular input as a spatial coordinate reference for multisensory processing.
• To explore how neuronal tuning properties influence multisensory integration and separation decisions.

Main Methods:

• Analysis of Macaque MST-d neuronal recordings.
• Construction and validation of a flexible Monte-Carlo probabilistic sampling (fMCS) model.
• Examination of the relationship between tuning response amplitude, tuning gradient, and decision-making.

Main Results:

• Visual input generally has a larger preferred tuning response than vestibular input in MST-d neurons.
• The fMCS model supports the hypothesis that visual cues serve as the primary coordinate reference for integration.
• Tuning gradient properties significantly affect integration/separation decisions, with dominant modalities benefiting from steep gradients.

Conclusions:

• MST-d neurons likely use a visual-based coordinate reference frame for integrating visual and vestibular motion cues.
• Both tuning response amplitude and tuning gradient play a joint role in determining the reference frame and modulating sensory integration.
• Understanding these mechanisms provides insight into how the brain constructs a coherent perception of self-motion.