Project Details
Description
Overview:
Page A
Effective memory storage is essential for human success in nearly every domain. Yet, few memories
are instantly embedded in the brain. Rather, the storage process evolves gradually. This gradual
evolution leading to secure memory storage reflects the operating principals of the brain
systems for storing factual and event knowledge (declarative memory). Understanding this pathway
to secure memory storage has been hampered by the difficulty of measuring the operative neural
signals. To solve this problem, this project integrates across multiple disciplines to non-invasively
track individual declarative memories as they compete with each other and grow stronger or
weaker during sleep. At the same time, the project addresses the Individuality/Variation Theme
by characterizing sources of variance in both reward processing and sleep replay dynamics
that determine which memory traces persist.
Intellectual Merit :
To understand how the fate of individual memories is determined during sleep, cutting-edge
developments in large-scale data analysis and machine learning will be leveraged along with
innovative experimental designs that provide temporal information about when a memory is reactivated
during sleep. Methodological advances produced through this project will thus open the door
to various other applications for deciphering codes of neural processing in the brain. Recall
is generally better for information that is ?tagged? at encoding as being important for the
future (e.g., due to a reward promised for successful remembering later), but there is no
such physical tag. Important information may benefit disproportionately because certain recently
encoded memories are replayed during sleep. A causal pathway to secure memory storage can
thus be described whereby importance at acquisition shapes neural dynamics during sleep by
selectively boosting the frequency and/or strength of memory replay; this differential replay
of important events can, in turn, give these memories a competitive advantage that boosts
subsequent recall. This project will test this hypothesis using an innovative paradigm for
targeted memory reactivation. Competition between memories is elicited during sleep through
a set of sound cues, each linked to two different objects. Multiple interlocking approaches
will track memory competition during sleep and how it shapes learning. Detailed predictions
will be derived from neural network modeling about how reward responses during wake shape
competitive dynamics during sleep, and how these competitive dynamics determine the eventual
fates of the competing memories. Predictions will be tested using fMRI to measure neural signals
of reward processing during encoding and EEG to measure brain activity during sleep with machine
learning methods to decode memory activation signals. Competitive dynamics during sleep will
be related to later memory and multivariate fMRI measures of representational change. The
project will provide, for the first time, a comprehensive look "under the hood" at the life
of a memory evolving during sleep and the sources of variance that influence memorability.
Pivotal knowledge will be gained about how variance in reward cues influences sleep replay
dynamics. Accordingly, the project will answer the big question of how variance in these dynamics
shape subsequent memory, greatly expanding current thinking about how memory works.
Broader Impacts :
Benefits encompass scientific understanding and methodological trailblazing, with significant
broader impa
Status | Finished |
---|---|
Effective start/end date | 9/1/15 → 2/28/19 |
Funding
- National Science Foundation (BCS-1533512)
Fingerprint
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.