Feature-Specific Salience Maps in Human Cortex

Area of Science:

• Neuroscience
• Cognitive Science
• Visual Perception

Background:

• Priority map theory posits that visual attention is guided by a priority map representing stimulus salience and task relevance.
• Computational models suggest separate feature dimensions (e.g., color, motion) contribute to this map.
• A key assumption is the existence of distinct neural regions representing these feature dimensions, which remains empirically untested.

Purpose of the Study:

• To test whether feature-selective visual cortex regions function as neural feature dimension maps.
• To investigate if these regions prioritize salient locations based on their preferred feature (color or motion).

Main Methods:

• Used functional magnetic resonance imaging (fMRI) to reconstruct spatial maps of brain activity.
• Participants viewed visual stimuli with salient regions defined by either color or motion.
• Analyzed activation patterns in retinotopic regions selective for color or motion.

Main Results:

• Reconstructed spatial maps showed activation localized to the salient stimulus position.
• Activation strength was highest in regions selective for the feature defining the salience (e.g., color-selective regions for color salience).
• Demonstrated a double dissociation: color regions prioritized color salience, motion regions prioritized motion salience.

Conclusions:

• Feature-selective extrastriate visual regions act as neural feature dimension maps.
• These regions index salient locations based on specific feature contrasts.
• Findings support the role of specialized cortical areas in guiding visual attention via feature-based salience computation.