Categorization training results in shape- and category-selective human neural plasticity

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Object category learning integrates bottom-up sensory data with top-down task demands.
• Cortical plasticity mechanisms are crucial for this integration and may be studied through category learning.
• A predictive model suggests task-independent shape representations facilitate subsequent learning.

Purpose of the Study:

• To investigate the neural mechanisms of object category learning.
• To test the hypothesis that category learning involves sharpening of shape-selective representations.
• To examine how training affects neural sensitivity to physical appearance versus category membership.

Main Methods:

• Functional magnetic resonance imaging (fMRI) with rapid-adaptation techniques.
• Training participants on a categorization task using morphed car stimuli.
• Analyzing neural responses in the lateral occipital cortex and lateral prefrontal cortex.

Main Results:

• Categorization training led to reduced adaptation for shape changes in the lateral occipital cortex, independent of category.
• A region in the lateral prefrontal cortex showed increased sensitivity to category membership post-training.
• Participants demonstrated improved discrimination performance for trained stimuli after categorization training.

Conclusions:

• Object category learning sharpens neural representations of physical shape in the visual cortex.
• Task-specific processing for category membership emerges in prefrontal cortex.
• These findings support a model where enhanced, task-independent representations facilitate learning and cognitive flexibility.