Predicting memory from the network structure of naturalistic events.

Area of Science:

• Cognitive Neuroscience
• Neuroimaging
• Memory Research

Background:

• Human memory recall involves not just individual events but also their interconnections.
• Existing research primarily examines discrete event memory, neglecting the impact of event network structure.
• Understanding how event network structure influences memory is crucial for comprehending complex experience recall.

Purpose of the Study:

• To investigate how the structural position of events within a network impacts their memory.
• To explore the neural mechanisms underlying the memory of central versus peripheral events.
• To determine if event centrality influences both memory encoding and retrieval processes.

Main Methods:

• Functional magnetic resonance imaging (fMRI) study with participants viewing and recalling realistic audiovisual narratives.
• Transforming narratives into event networks to quantify event centrality based on semantic/causal connections.
• Analyzing hippocampal and cortical activation patterns during memory encoding and retrieval.

Main Results:

• Events with higher centrality (stronger connections) were better remembered.
• Central events elicited greater hippocampal event boundary responses during encoding.
• High centrality correlated with increased activation in episodic recollection areas during recall and more consistent neural representations across individuals.

Conclusions:

• Memory recall and representation are significantly shaped by an event's position within a network.
• Central events are prioritized in memory due to enhanced neural processing during encoding and retrieval.
• The network structure of real-world experiences influences the reliability and accessibility of episodic memory.