Awake ripples enhance emotional memory encoding in the human brain

The study investigates the neural mechanisms underlying the enhanced memory for emotional experiences. Prior research suggests that this advantage stems from the interplay between the amygdala and hippocampus during memory consolidation, with neuromodulatory effects influencing plasticity. Memory reinstatement during the immediate post-encoding period has also been shown to predict subsequent memory performance. This study focuses on ripples, transient hippocampal oscillations (80-150 Hz), known for their involvement in binding distributed memory traces and behaviorally relevant reactivation of emotional memories. The hypothesis is that post-encoding ripples facilitate emotional memory discrimination via coordinated hippocampal-amygdala memory reinstatement or by maintaining stimulus information in working memory. This process would manifest as increased stimulus similarity during post-encoding ripples. Intracranial EEG recordings from epilepsy patients performing an emotional encoding and discrimination task were used to test this hypothesis. The study builds upon previous findings demonstrating better memory discrimination for arousing stimuli and the role of ripples in memory processing.

The introduction section extensively cites previous research supporting the prioritized encoding of emotional experiences, highlighting the involvement of neuromodulatory effects and amygdala-hippocampal interactions. Studies linking memory reinstatement during the immediate post-encoding period to later memory performance are also reviewed. The role of ripples in the binding of memory traces and their association with reactivation of emotional memory are discussed, citing relevant literature. This review provides a comprehensive background on the existing knowledge about emotional memory and the specific role of hippocampal ripples, setting the stage for the study's specific hypotheses and methodology.

Seven epilepsy patients with intracranial EEG electrodes implanted in the amygdala and hippocampus participated in an emotional memory encoding and discrimination task. During encoding, participants viewed images and rated their emotional valence (negative, neutral, positive). During retrieval, they were presented with repeat, lure (similar), or novel images and classified them as old or new. Memory discrimination accuracy was assessed. Post-encoding ripples were detected in the hippocampal EEG using established methods, and their rate was analyzed in relation to stimulus arousal and memory accuracy. Representational Similarity Analysis (RSA) was employed to quantify stimulus similarity in the HFA (30-280 Hz) power spectral vectors, examining similarity around ripple peaks in both amygdala and hippocampus. Mutual information analysis was used to determine the directional influence between amygdala and hippocampal activity. Statistical analyses included linear mixed-effects models, Wilcoxon signed-rank tests, ANOVAs, logistic regressions, and non-parametric cluster-based permutation tests. Detailed signal preprocessing steps, including artifact removal and filtering, are described. Electrode localization methods and the specifics of the EEMD (Ensemble Empirical Mode Decomposition) used for HFA analysis are also outlined. The behavioral aspects of the study, such as stimulus selection and rating procedures, are thoroughly described.

The study found that participants showed enhanced memory discrimination for arousing stimuli. Post-encoding ripple rate was positively correlated with both stimulus-induced arousal and correct lure discrimination. This association was specific to the post-encoding period. There was a positive association between post-encoding ripple rate and stimulus-induced arousal. Post-encoding ripple rate predicted correct lure discrimination during retrieval. RSA revealed that stimulus similarity during post-encoding ripples was higher for arousing and correctly discriminated stimuli. This similarity showed region-specific timing, with amygdala similarity preceding hippocampal similarity. Amygdala ripple-locked similarity correlated more strongly with stimulus arousal, while hippocampal similarity correlated more with correct lure discrimination, demonstrating a double dissociation. Joint analysis revealed a significant increase in ripple-locked stimulus similarity in both amygdala and hippocampus only for correctly discriminated stimuli, with the amygdala leading the hippocampus. Mutual information analysis confirmed a unidirectional influence from amygdala to hippocampus before the ripple peak. These findings suggest a coordinated process where amygdala activity triggers hippocampal ripples and enhances emotional memory.

The study's findings provide electrophysiological evidence supporting the hypothesis that post-encoding ripples enhance emotional memory. The observed association between ripple rate, arousal, and memory accuracy directly links neural activity to behavioral performance. The region-specific timing of stimulus similarity in the amygdala and hippocampus suggests a sequential process, with amygdala-mediated emotional tagging preceding hippocampal consolidation. The coordinated increase in stimulus similarity across regions during ripples suggests a mechanism for binding emotional and contextual aspects of memory. The study supports the idea that ripples play a crucial role in memory consolidation and emotional memory enhancement, particularly during the immediate post-encoding period. The results extend previous work on the role of ripples in memory by demonstrating their involvement in emotional memory processing specifically.

This study provides compelling evidence that post-encoding hippocampal ripples, particularly those showing increased stimulus similarity, are a critical neural mechanism for enhanced emotional memory encoding in humans. The findings highlight the interplay between the amygdala and hippocampus in this process. Future studies could explore the specific cellular and molecular mechanisms underlying ripple-mediated memory enhancement and investigate the role of different types of emotional stimuli. Examining individual differences in ripple activity and their relationship to emotional memory performance could also be a promising avenue for future research.

The relatively small sample size (7 participants) due to the scarcity of patients with simultaneous amygdala and hippocampal recordings outside seizure zones is a significant limitation. While the effects were consistent across individual participants, larger studies are needed to confirm the findings and establish the robustness of the observed relationships. The study focused on lure discrimination, and it is not certain whether the findings generalize to other forms of memory recognition. The use of epilepsy patients might affect the results as this might not generalize to the healthy population.