A computational predictor of human episodic memory based on a theta phase precession network

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Theta phase precession, observed in rodents, links place cell firing to the theta rhythm and is crucial for memory formation via spike-time-dependent plasticity (STDP).
• The existence and role of theta phase precession in primates, including humans, remain unclear, despite its proposed involvement in object-place association memory.

Purpose of the Study:

• To evaluate the presence and significance of theta phase precession in human memory recall.
• To determine if theta phase precession dynamics can predict human memory performance using eye movement and EEG data.

Main Methods:

• A theory-experiment combined analysis was employed, simulating a human hippocampal network using theta phase precession dynamics.
• Human eye movement and electroencephalogram (EEG) data during memory encoding were analyzed.
• Computational recall from the simulated network was compared with human memory recall performance.

Main Results:

• The computational recall, based on theta phase precession dynamics, showed a significant correlation with human memory recall performance.
• Computational predictors lacking theta phase precession did not significantly correlate with subsequent memory recall.
• The predictive power of theta phase precession exceeded that of traditional experimental predictors.

Conclusions:

• Theta phase precession dynamics are essential for accurately predicting human recall performance from eye movement and EEG data.
• This neural dynamic effectively extracts memory-dependent components from spatio-temporal data, functioning as an associative network.
• Theta phase precession may represent a conserved neural mechanism for environmental memory formation in both rodents and humans.