Neural similarity between overlapping events at learning differentially affects reinstatement across the cortex

Area of Science:

• Neuroscience
• Cognitive Psychology
• Memory Research

Background:

• Episodic memory often involves highly similar events with overlapping elements.
• The brain must balance differentiating similar memories to prevent interference and integrating them to link shared information.
• Neural mechanisms supporting these dual functions of differentiation and integration remain unclear.

Purpose of the Study:

• To investigate how the brain encodes highly overlapping naturalistic events.
• To examine the effects of neural differentiation versus integration during encoding on later memory retrieval.
• To understand the distinct roles of cortical regions in supporting these processes.

Main Methods:

• Utilized multivoxel pattern similarity analysis (MVPA) on fMRI data.
• Employed neural-network analysis to assess visual similarity of stimuli.
• Analyzed neural activity patterns during learning and recall of naturalistic video stimuli.

Main Results:

• Overlapping neural activity patterns in temporal, parietal, and occipital regions indicated memory integration for visually similar videos.
• Differentiation during encoding predicted later neural reinstatement in occipital visual processing regions.
• Integration in temporal and parietal higher-level sensory regions predicted greater reinstatement, accuracy, and vividness at recall.

Conclusions:

• Neural encoding of similar episodic memories involves both differentiation and integration across different cortical areas.
• These processes have divergent impacts on memory recall, with integration in higher-level sensory regions being particularly beneficial for accurate and vivid recall.