Attention Modulates Stimulus Representations in Neural Feature Dimension Maps

Area of Science:

• Neuroscience
• Cognitive Science
• Visual Perception

Background:

• Computational models suggest attention relies on dynamically weighted spatial feature maps.
• Prior fMRI studies show enhanced responses in feature-specific regions when those features are attended.

Purpose of the Study:

• To investigate if multivariate activation patterns form spatial "feature dimension maps" guiding attentional priority.
• To test if these maps are dynamically reweighted based on task demands.

Main Methods:

• Used functional magnetic resonance imaging (fMRI) to record brain activity.
• Reconstructed spatial maps from fMRI data using a spatial inverted encoding model.
• Participants performed a feature-selective attention task involving stimuli with color and motion.

Main Results:

• Spatial maps showed selective enhancement in color-selective regions when color was attended.
• Spatial maps showed enhancement in motion-selective regions when motion was attended.
• Enhancement was localized in color regions but more global in motion regions, depending on task relevance.

Conclusions:

• Feature-selective cortical regions support "neural feature dimension maps" for different visual features.
• These maps are dynamically reweighted by task demands to guide attention.
• This mechanism helps prioritize relevant locations based on important features for visual behavior.