Decoding semantic categories: insights from an fMRI ALE meta analysis.

Area of Science:

• Neuroscience
• Cognitive Science
• Neuroimaging

Background:

• The human brain organizes conceptual knowledge into semantic categories.
• Understanding the neural basis of semantic categories and their representations is crucial.
• Existing research lacks clarity on the extent of shared versus distinct neural representations across semantic domains.

Purpose of the Study:

• To clarify the organizational structure of semantic categories in the brain.
• To identify consistent, modality-controlled activation patterns across key semantic domains using fMRI.
• To quantify the distinctiveness and overlap of neural patterns for applications like semantic brain-computer interfaces (BCI).

Main Methods:

• Systematic review and meta-analysis of 75 fMRI studies following PRISMA guidelines.
• Analysis covered six semantic categories: animals, tools, food, music, body parts, and pain.
• Activation Likelihood Estimation (ALE) and Jaccard-based overlap analyses were used to identify and quantify activation patterns and their overlap across modalities.

Main Results:

• Distinct yet partially overlapping activation networks were identified for each semantic category.
• Tools and animals shared activity in visual processing regions; food activated limbic/subcortical areas.
• Music and animal sounds overlapped in auditory cortices; body parts and pain engaged specific occipito-parietal and cingulo-insular networks, respectively, revealing a hierarchical and modality-dependent architecture.

Conclusions:

• Semantic knowledge is distributed and differentiated across cortical systems in a hierarchical and modality-dependent manner.
• Conceptual content and sensory modality jointly shape neural organization.
• Findings refine models of semantic cognition and provide a basis for semantic neural decoding and BCI development.