Brain activity across the development of automatic categorization: a comparison of categorization tasks using...

Area of Science:

• Cognitive Neuroscience
• Neuroimaging

Background:

• Initial learning of rule-based and information-integration categorization tasks involves distinct neural networks.
• With increased automaticity, categorization behavior and associated neural activity converge, primarily involving cortical regions.

Purpose of the Study:

• To directly compare the development of automaticity across different categorization tasks using multi-voxel pattern analysis.
• To investigate how neural activation patterns and representational similarities evolve with extensive training in distinct categorization tasks.

Main Methods:

• Employed multi-voxel pattern analysis (MVPA) on functional magnetic resonance imaging (fMRI) data.
• Trained three participant groups on distinct categorization tasks (information-integration or rule-based) across four sessions within an MRI scanner.
• Conducted common patterns, unique patterns, and representational similarity analyses on regions of interest (ROIs).

Main Results:

• Common activation patterns were observed in motor areas and basal ganglia early in training, persisting only in motor areas later.
• Unique activation patterns in cortical and subcortical areas diminished significantly with training.
• Representational similarity between brain regions increased across tasks as automaticity developed.

Conclusions:

• The development of automaticity in categorization tasks leads to increasingly similar neural representations across different task types.
• Motor areas and basal ganglia play a role in early learning, with a shift towards cortical dominance for automatic categorization.
• Neural processes underlying categorization become more unified with extensive practice, regardless of the initial task structure.