Hippocampal-Medial Prefrontal Event Segmentation and Integration Contribute to Episodic Memory Formation

How do we encode our continuous life experiences for later retrieval? Theories of event segmentation and integration suggest that the hippocampus binds separately represented events into an ordered narrative. Using a functional Magnetic Resonance Imaging (fMRI) movie watching-recall dataset, we quantified two types of neural similarities (i.e., activation pattern similarity and within-region voxel-based connectivity pattern similarity) between separate events during movie watching and related them to subsequent retrieval of events as well as retrieval of sequential order. We demonstrated that compared with forgotten events, successfully remembered events were associated with distinct activation patterns in the hippocampus and medial prefrontal cortex. In contrast, similar connectivity pattern between events were associated with memory formation and were also relevant for retaining events in the correct order. We applied the same approaches to an independent movie watching fMRI dataset as validation and highlighted again the role of hippocampal activation pattern and connectivity pattern in memory formation. We propose that distinct activation patterns represent neural segmentation of events, while similar connectivity patterns encode context information and, therefore, integrate events into a narrative. Our results provide novel evidence for the role of hippocampal-medial prefrontal event segmentation and integration in episodic memory formation of real-life experience.

Automated summary

Research indicates that successfully remembered events are associated with distinct activation patterns in the hippocampus and medial prefrontal cortex, while similar connectivity patterns between events are linked to memory formation and retention of order. Different activation patterns represent neural segmentation of events, while similar connectivity patterns encode context information to integrate events into a narrative.