Do different brain rhythms handle seeing versus linking a memory?
Drops in cortical alpha/beta rhythms predicted later memory while people were taking in information, whereas hippocampal theta-gamma coupling predicted memory only when they were binding the pieces together.
Source
Disentangling neocortical alpha/beta and hippocampal theta/gamma oscillations in human episodic memory formation
Study at a glance
- Design
- Human experiment — Within-subject MEG study of an associative memory task: participants saw an object, a pattern and a scene, then had a short window to bind them; brain activity in each phase was related to how many associates were later recalled.
- N
- N=17 · 28 recruited; participants were excluded for head movement (1), poor data quality (4) or extreme memory performance (6), leaving 17 for analysis.
- Population
- Healthy young adults (analysed sample aged 20 to 32) at the University of Birmingham
- Outcome
- Sensor and source-level alpha/beta power and hippocampal theta-gamma phase-amplitude coupling as predictors of the number of associates later recalled
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Seventeen adults in an MEG scanner saw sequences of a line-drawn object, a coloured pattern and a scene, and then had a short window to combine them into one vivid mental image. After a short distractor task, they were shown the object and had to pick its pattern and scene from three options each. The researchers used single-trial regressions to ask how brain rhythms during the perception phase and the binding phase predicted how many associates were later remembered.
What they found
During perception of the sequence, lower alpha/beta power (roughly 8 to 15 Hz) over posterior sensors, localised to early visual cortex, predicted more items recalled, but this was not seen during binding. During the binding window, stronger coupling between hippocampal theta phase and gamma amplitude predicted more items recalled, but this was not seen during perception. A formal interaction test confirmed a double dissociation between the two stages. Memory was better for scenes than for patterns.
The limits
What it doesn't show
The analysed sample was small (17 of 28 recruited) after many exclusions, including six for extreme memory performance. Measuring the deep hippocampus with MEG relies on source reconstruction, and the authors caution that apparent left-right differences may be artefacts. Perception and binding cannot be completely separated in time, and the alpha effect could reflect attention rather than information representation. Eye movements were not recorded, so ocular activity could partly mediate the effects.
Key terms
- Episodic memory
- Memory for a specific personal event anchored to a particular time and place.
- Alpha/beta desynchronisation
- A drop in 8 to 30 Hz oscillatory power, thought to reflect less synchronised, more information-rich cortical activity.
- Theta-gamma phase-amplitude coupling
- When the strength of fast gamma oscillations rises and falls with the phase of slower theta oscillations, proposed to support synaptic plasticity for binding.
- Subsequent memory effect
- A difference in brain activity during learning that predicts which items will later be remembered.
- Magnetoencephalography (MEG)
- A method that records the magnetic fields produced by neural currents, giving millisecond timing of brain activity.
Flashcards
0 of 12 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
During which phase did lower alpha/beta power predict better memory?
Common questions
Why would less synchronised cortical activity help memory?
Information theory suggests unpredictable, desynchronised activity can carry more information, so lower alpha/beta power may mean richer cortical representation of the incoming stimulus.
Can MEG really measure the hippocampus?
It is harder than for surface cortex, but the authors used source reconstruction, checked that other lobes did not show the same coupling, and cite studies validating MEG hippocampal signals against intracranial recordings.
Why was there no memory-related change in gamma power alone during binding?
If gamma increases only at particular theta phases and theta is not locked to the stimulus, averaging across trials cancels the gamma effect, while coupling measures still reveal it.
More on Neural oscillations