Predictive feedback to V1 dynamically updates with sensory input.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Visual Perception

Background:

• Predictive coding theories suggest the brain uses internal models to anticipate sensory input.
• Hierarchical models in vision propose feedback from higher areas to early visual cortex (V1).
• Mismatch between predictions and sensory input in V1 typically amplifies neural responses.

Purpose of the Study:

• To investigate if the brain's predictive feedback mechanism dynamically updates its retinotopic location following eye movements.
• To test the role of eye movements in predictive coding within the visual system.
• To examine how apparent motion illusions influence predictive feedback in V1.

Main Methods:

• Utilized functional brain imaging (fMRI) and eye-tracking technology.
• Presented an apparent motion illusion to induce internal motion models and predictive feedback.
• Measured brain activity in V1 before and after saccadic eye movements.

Main Results:

• Observed attenuated Blood-Oxygen-Level-Dependent (BOLD) responses in V1 for predicted stimuli at new post-saccadic locations.
• Demonstrated that predictions are updated to match the new retinotopic location after an eye movement.
• Showed that the brain anticipates sensory input at the upcoming location.

Conclusions:

• Pre-saccadic predictions are updated in their retinotopic location to anticipate post-saccadic visual input.
• This finding validates dynamic predictive coding theories in V1, highlighting the brain's adaptive predictive capabilities.
• The visual system actively adjusts its internal models in real-time to account for changes in visual field location due to eye movements.