Learning to integrate contradictory multisensory self-motion cue pairings.

Area of Science:

• Neuroscience
• Perception
• Human Multisensory Integration

Background:

• Humans integrate multisensory information to reduce perceptual uncertainty.
• Integration fails if sensory signals lack a common causal origin (spatiotemporal discrepancies).
• Visual and vestibular cues for self-motion are integrated optimally, but limits of cue conflict are unknown.

Purpose of the Study:

• To investigate if humans can learn to integrate arbitrary, co-occurring visual and vestibular self-motion cues.
• To determine if integration occurs even with different rotation axes (yaw vs. roll/pitch).

Main Methods:

• Participants compared sizes of successive whole-body rotations using visual, vestibular, or combined stimuli.
• Vestibular stimulus: yaw rotation; Visual stimulus: roll (Exp 1) or pitch (Exp 2) rotation.
• Bimodal thresholds were compared to predictions from single-cue thresholds.

Main Results:

• Human subjects optimally integrated vestibular (yaw) and visual (roll/pitch) self-motion cues.
• Integration occurred despite the cues signaling rotation around different axes.
• Results suggest inference of a common cause for temporally co-occurring, spatially distinct self-motion cues.

Conclusions:

• The human brain can learn to integrate multisensory self-motion information even with conflicting spatial cues.
• This integration implies inferring a common cause for disparate sensory inputs.
• Findings have implications for understanding bodily illusions and cross-modal adaptation.