Perceptual Priors Update Contextual Feedback Processing in V1

Area of Science:

• Neuroscience
• Cognitive Science
• Visual Perception

Background:

• Contextual information and prior knowledge significantly influence perceptual processing, enabling recognition of degraded visual inputs.
• Neuronal processing of identical sensory data varies with context, aligning with predictive processing models where higher brain areas generate predictions for sensory interpretation.
• Cortical feedback signals transmit predictions to sensory areas, suggesting knowledge acquisition refines internal models and improves sensory estimation accuracy.

Purpose of the Study:

• To investigate how acquired knowledge modulates contextual feedback processing in the primary visual cortex (V1).
• To determine if general category knowledge or specific image information impacts feedback signals in response to ambiguous visual stimuli.
• To explore the role of internal models in resolving sensory ambiguities through top-down modulation.

Main Methods:

• Utilized Mooney images, ambiguous two-tone stimuli, to assess recognition thresholds without prior knowledge.
• Conducted two behavioral experiments and one functional magnetic resonance imaging (fMRI) study.
• Participants acquired either general category or specific information about Mooney images; fMRI used partially occluded images to examine feedback modulation in V1.

Main Results:

• General knowledge of image categories was sufficient to enhance the recognition of ambiguous Mooney images.
• Perceptual priors incorporating image-specific information modulated contextual feedback processing in early visual areas.
• Previously ambiguous images showed altered feedback signal processing after knowledge acquisition.

Conclusions:

• Acquiring knowledge, even general category information, improves the ability to recognize ambiguous visual stimuli.
• Specific perceptual priors dynamically modulate feedback mechanisms in early visual cortex.
• Predictive processing models are supported by evidence showing knowledge-driven refinement of feedback signals in visual perception.