Decoding Modality-Independent Emotional Representations in Language-Related Brain Networks : An fMRI Searchlight MVPA Study

Sammanfattning

The present thesis investigates how emotion-related information is represented across the brain, whether these representations depend on stimulus sensory modality, and whether language-related regions support modality-independent encoding of emotional categories. The central research question is whether language-related regions preferentially exhibit stronger modality-independent representations of emotional categories than sensory regions. To address this question, the study applies multivariate pattern analysis (MVPA) to decode emotional categories from spatially distributed patterns of neuronal activity, measured using functional Magnetic Resonance Imaging (fMRI). The findings suggest that emotional representations are largely shaped by sensory modality and task demands, with only limited evidence for fully abstract, modality-independent processing in language-related regions, which appears to be restricted to certain emotion pairs and is not consistent across tasks. Overall, the study provides a foundation for future research on the relationship between emotion categorization, language, and cognition.