Decoding the emotional modulation of memory encoding: a multivariate fmri study of arousal and valence

Emotional events are often remembered better than neutral ones, yet the distinct contributions of valence (positive/negative) and arousal (calm/exciting) to the neural patterns supporting memory encoding are not fully understood. We leveraged multivariate pattern analysis (MVPA) to dissociate these influences. We analysed fMRI data from a large cohort of participants during an emotional picture encoding task with a subsequent free recall test. A whole-brain searchlight Support Vector Machine (SVM) pipeline was implemented to decode subsequent memory success ('recalled' vs. 'not recalled') separately within high, medium, and low arousal contexts, as well as within positive, neutral, and negative valence contexts. Individual-level statistical validation was performed using permutation testing and False Discovery Rate (FDR) correction, followed by a group-level analysis to identify consistent neural patterns. Behaviourally, we found a significant main effect of arousal on recall performance, but no significant effect of valence. Neurally, our decoding results mirrored this pattern. The classification of subsequent memory was significantly more accurate and robust within high-arousal contexts compared to lowarousal contexts, with key informative regions localised to the ventral visual stream and temporal memory systems, including the inferior temporal and fusiform gyri. In contrast, while we identified distinct neural patterns for memory encoding under different valence conditions, their decoding accuracies did not significantly differ. These findings provide powerful multivariate evidence that emotional arousal, rather than valence, is the primary modulator of the neural fidelity of memory encoding. Our results suggest that high arousal enhances memory by sharpening the distinctiveness of neural representations within the visual-temporal network, demonstrating the utility of MVPA in providing finer-grained insights into the mechanisms of emotional memory.

Author dc:creator

• Yuan, Yuan

Advisor dc:contributor.advisor

• Luksys, Gedi

• Emotional Modulation
• Memory Encoding
• Arousal
• Valence
• Multivariate Pattern Analysis (MVPA)

Language dc:language.iso

en

Identifier URI

https://doi.org/10.7488/era/6984

OAI identifier oai:identifier

oai:era.ed.ac.uk:1842/44467