Neural mechanisms for discounting head-roll-induced retinal motion

Area of Science:

• Neuroscience
• Human Vestibular System
• Multisensory Integration

Background:

• Studies in macaques identified neurons in dorsal medial superior temporal area (MSTd), ventral intraparietal area (VIP), and visual posterior sylvian area (VPS) tuned to self-motion direction across visual and vestibular modalities.
• Some neurons integrate congruent sensory signals for trajectory encoding, while others use opposing signals to discount irrelevant head motion.
• The existence of a similar system in humans remained unknown.

Purpose of the Study:

• To investigate whether human multisensory cortical regions exhibit neural populations tuned to self-motion direction, similar to findings in macaques.
• To explore the integration and potential discounting of visual and vestibular self-motion cues in the human brain.

Main Methods:

• Artificial vestibular stimulation was applied to human participants during functional magnetic resonance imaging (fMRI).
• Visual stimulation was carefully calibrated to emulate congruent or opposite cue conditions relative to vestibular input.
• Multivoxel pattern analysis (MVPA) was used to decode stimulus phase from fMRI data in specific brain regions.

Main Results:

• MVPA successfully classified stimulus phase in human MST and putative VIP, indicating separate neural populations responding to congruent and opposite cue combinations.
• Decoding of stimulus phase was also achieved near the parieto-insular vestibular cortex, potentially involving a human homolog of macaque VPS.
• These findings suggest the presence of distinct neural mechanisms for processing integrated and conflicting self-motion cues in humans.

Conclusions:

• Human MST and VIP contain neural populations that process self-motion information from combined visual and vestibular inputs.
• The human brain appears to employ mechanisms for both integrating congruent self-motion cues and potentially discounting conflicting ones.
• This study provides evidence for a human neural system analogous to the macaque system for encoding self-motion, supporting multisensory integration theories.