Separability of abstract-category and specific-exemplar visual object subsystems: evidence from fMRI pattern analysis

Area of Science:

• Neuroscience
• Cognitive Psychology
• Neuroimaging

Background:

• Neural systems for abstract-category (AC) and specific-exemplar (SE) object recognition are thought to be dissociable.
• The degree of separability between these systems is crucial for evaluating relevant theories of object recognition.

Purpose of the Study:

• To investigate the neural representations of visual object categories and exemplars.
• To assess the separability of abstract-category (AC) and specific-exemplar (SE) object processing subsystems using multivariate pattern analysis (MVPA).

Main Methods:

• Utilized multi-voxel pattern analyses (MVPA) of functional magnetic resonance imaging (fMRI) data during a repetition-priming task.
• Trained classifiers to distinguish between same-exemplar primed and word-primed objects in visual object processing areas.
• Assessed classifier performance for discriminating different-exemplar primed objects (AC priming) and same-exemplar primed objects (SE priming).

Main Results:

• Repetition priming in occipital-temporal regions showed asymmetric organization.
• Abstract-category (AC) priming was more prevalent in the left hemisphere.
• Specific-exemplar (SE) priming was more prevalent in the right hemisphere.

Conclusions:

• The neural subsystems for AC and SE object recognition are weakly modular.
• These findings suggest a degree of specialization within visual object processing, rather than a unified system.