Neuronal tuning and population representations of shape and category in human visual cortex

Area of Science:

• Neuroscience
• Cognitive Science
• Visual Perception

Background:

• Object recognition and categorization are crucial cognitive functions.
• These processes heavily involve the human ventral visual stream.
• Neuronal tuning for object shape and category in this area remains poorly understood.

Purpose of the Study:

• To investigate neuronal tuning properties for object shape and category in the human ventral visual stream.
• To dissociate shape from category dimensions in stimuli.
• To reconcile findings between human imaging and nonhuman primate single-cell studies.

Main Methods:

• Large-scale recordings of spiking activity in the human Lateral Occipital Complex (LOC).
• Presentation of stimuli with dissociated shape and category dimensions.
• Analysis of neuronal responses using linear decoders.

Main Results:

• Neuronal representations in the LOC were predominantly shape-based.
• Category-like encoding was observed, particularly for animal images.
• Linear decoders could classify stimulus category even from shape-based data.
• Significant interactions between shape and category tuning were detected.

Conclusions:

• Human ventral stream neurons encode object information based primarily on shape.
• Both shape and category information are represented, with evidence of interaction.
• These findings provide crucial insights into neuronal-level object coding in humans.