Dissociating the Neural Representation of Scene Boundary and Extended Surfaces in Human Visual Scene Processing

Area of Science:

• Neuroscience
• Cognitive Science
• Environmental Psychology

Background:

• Humans use environmental boundaries for navigation and place recognition.
• The precise features defining effective cortical boundaries for scene processing are not fully understood.
• While walls are common boundaries, their ability to mark 3D shape may be key.

Purpose of the Study:

• To investigate whether the human brain encodes environmental boundaries based on 3D shape rather than surface continuity.
• To test if geometrically equivalent boundaries, irrespective of surface type (walls vs. poles), are processed similarly.
• To explore the neural basis of boundary representation in scene-selective cortical regions.

Main Methods:

• Used functional magnetic resonance imaging (fMRI) to measure brain activity.
• Presented participants with artificial images of boundaries defining identical 3D shapes.
• Systematically varied boundary structure by using walls versus a varying number of poles.

Main Results:

• The parahippocampal place area showed a 'wall-like' representation even with isolated poles, indicating sensitivity to non-wall boundaries and environmental shape.
• The occipital place area responded to variations in boundary structure, reflecting sensitivity to continuous, surface-like properties.
• Distinct cortical regions process different boundary features crucial for encoding environmental geometry.

Conclusions:

• The human scene-processing system is sensitive to environmental shape defined by non-wall boundaries.
• Neural representations of environmental geometry depend on distinct boundary features processed in different brain regions.
• This research clarifies how the brain encodes spatial environments using diverse boundary types.