Phase separation of competing memories along the human hippocampal theta rhythm
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
This pre-registration study by Kerrén et al. used a proactive interference task in combination with MEG recordings on humans to test predictions of a previous computational model postulating that neural representations of competing memories are associated with varied phases of the hippocampus theta-band rhythm. Their results largely confirmed the hypothesis and suggest that reactivations of target and competitor memories indeed occur at different phases of theta oscillations, which is further related to the intrusion effect in behavior.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #2 agreed to share their name with the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Neuroscience (eLife)
Abstract
Competition between overlapping memories is considered one of the major causes of forgetting, and it is still unknown how the human brain resolves such mnemonic conflict. In the present magnetoencephalography (MEG) study, we empirically tested a computational model that leverages an oscillating inhibition algorithm to minimise overlap between memories. We used a proactive interference task, where a reminder word could be associated with either a single image (non-competitive condition) or two competing images, and participants were asked to always recall the most recently learned word–image association. Time-resolved pattern classifiers were trained to detect the reactivated content of target and competitor memories from MEG sensor patterns, and the timing of these neural reactivations was analysed relative to the phase of the dominant hippocampal 3 Hz theta oscillation. In line with our pre-registered hypotheses, target and competitor reactivations locked to different phases of the hippocampal theta rhythm after several repeated recalls. Participants who behaviourally experienced lower levels of interference also showed larger phase separation between the two overlapping memories. The findings provide evidence that the temporal segregation of memories, orchestrated by slow oscillations, plays a functional role in resolving mnemonic competition by separating and prioritising relevant memories under conditions of high interference.
Article activity feed
-
-
-
Evaluation Summary:
This pre-registration study by Kerrén et al. used a proactive interference task in combination with MEG recordings on humans to test predictions of a previous computational model postulating that neural representations of competing memories are associated with varied phases of the hippocampus theta-band rhythm. Their results largely confirmed the hypothesis and suggest that reactivations of target and competitor memories indeed occur at different phases of theta oscillations, which is further related to the intrusion effect in behavior.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #2 agreed to share their name with the authors.)
-
Reviewer #1 (Public Review):
This pre-registration study by Kerrén et al. examined the hypothesis that the brain resolves competition between overlapping memories based on phase separation of hippocampus theta oscillations. By applying a time-resolved decoding analysis, they demonstrate that the reactivations of target and competitor memories are locked to the varied phase of theta-band oscillation after repeated recalls. Moreover, subjects with larger phase separation show less memory interference. The study provides new evidence supporting the phase-coding neural mechanism to alleviate memory interference of multiple items.
Overall, this is a very interesting report testing an influential oscillatory-phase-based hypothesis in the memory field and would bring broad impacts to other fields, such as perception, attention, and decision …
Reviewer #1 (Public Review):
This pre-registration study by Kerrén et al. examined the hypothesis that the brain resolves competition between overlapping memories based on phase separation of hippocampus theta oscillations. By applying a time-resolved decoding analysis, they demonstrate that the reactivations of target and competitor memories are locked to the varied phase of theta-band oscillation after repeated recalls. Moreover, subjects with larger phase separation show less memory interference. The study provides new evidence supporting the phase-coding neural mechanism to alleviate memory interference of multiple items.
Overall, this is a very interesting report testing an influential oscillatory-phase-based hypothesis in the memory field and would bring broad impacts to other fields, such as perception, attention, and decision making, given that lessening inference of distractor to target is a fundamental challenge.
Meanwhile, several aspects of the results need more evidence to strengthen the conclusion. The major weakness is the lack of significant decoding for the target and competitor by themselves, although I understand that the main hypothesis focuses on their different phase-locking relationship. Meanwhile, less significant decoding performance, I believe, is very crucial to verify the LDA analysis and data quality. Similarly, the original decoding performance time course did not show a clear out-of-phase pattern as revealed in the phase analysis. Overall, the authors need more results to confirm that the phase separation results are based on genuine reactivations and out-of-phase relationships.
-
Reviewer #2 (Public Review):
This manuscript describes the results of an empirical study designed to test the specific predictions of the computational model described in Norman et al., 2006 and, more generally, to better understand the relationship between oscillatory dynamics and memory. In particular, the authors seek to advance the mechanistic understanding of how competition during mnemonic retrieval changes the accessibility of the memory traces involved. Toward this end, the described experiment examines the brain activity, as recorded with MEG, of human participants as they study lists of word-image pairs. Critically, individual words were paired with either one or two different images creating conditions where retrieval of the most recent pairing would either generate minimal competition (case of one paired image) or elevated …
Reviewer #2 (Public Review):
This manuscript describes the results of an empirical study designed to test the specific predictions of the computational model described in Norman et al., 2006 and, more generally, to better understand the relationship between oscillatory dynamics and memory. In particular, the authors seek to advance the mechanistic understanding of how competition during mnemonic retrieval changes the accessibility of the memory traces involved. Toward this end, the described experiment examines the brain activity, as recorded with MEG, of human participants as they study lists of word-image pairs. Critically, individual words were paired with either one or two different images creating conditions where retrieval of the most recent pairing would either generate minimal competition (case of one paired image) or elevated competition (case of two paired images).
The authors partially confirm a set of specific and rigorous preregistered predictions regarding the differences between these conditions, including increased frontal theta on competition trials relative to minimal competition trials, increased evidence of reactivation of the competing memory on competition trials, and, most relevant for the predictions of the Norman et al., 2006 model, reactivation of target and competitor memories occurred at different phases of a 3Hz oscillation and were functionally related to the intrusion rate of the competing memories.
Discrepancies between the actual results and the preregistered predictions were consistently, by my interpretation, non-substantial. For example, the frontal theta was modestly displaced laterally. Also, activation of the competitor was only detectable on the third repetition but not earlier repetitions. Certainly, additional work will help bring light to these differences, but overall I was impressed with the overall quality of the alignment between the predictions and the outcomes.
Methodologically, I believe this paper brings innovation and rigor to the multivariate pattern analysis (MVPA) based exploration of brain activity. Most powerfully in this context was the use of separate super-ordinate and sub-ordinate category classifiers. In doing so, testing for 'competitor reactivation' (e.g., reactivation of a previously associated scene when the participant was retrieving the more recently associated object image) was not simply looking for 'scene reactivation.' This is important because 'scene reactivation' is confounded with what might be called 'not object activation.' Rather, the authors make clever use of a classifier constructed to distinguish between subordinate categories of scene (e.g., outdoor vs. indoor). For this classifier to generate meaningful output and for that output to reflect the reactivation of the competitor, the pattern of brain activity must be specific to scenes of the right sort. Without this operationalization of the methods, the results of this work would have been difficult to interpret. As it is, this work provides a strong body of evidence regarding the relationship between oscillatory dynamics and memory.
Scientifically, this manuscript is likely to be relevant to a broad range of scientists studying both human memory and the neurobiology of learning and memory. The results provide a portrait of the neural dynamics underlying competition-dependent memory that parallels the types provided by animal in-vivo recordings. The fact that the results are tied to a computational model described by Norman et al. 2006 makes the current results all the more interpretable.
-
Reviewer #3 (Public Review):
This manuscript uses MEG data acquired from human participants to examine whether representations of competing memories are associated with different phases of the theta rhythm in the human hippocampus. In brief, the authors use a proactive interference task in which subjects are asked to associate a word with two competing images and then subsequently recall the most recent image. Using pattern classifiers on the MEG data, the authors are able to decode reactivated content of the target and competitor memories and find that these patterns appear locked to different phases of the hippocampal rhythm. They also show that those subjects with worse memory performance had fewer differences in the phases to which target and competitor memories are locked. Together, the data provide support for a computational …
Reviewer #3 (Public Review):
This manuscript uses MEG data acquired from human participants to examine whether representations of competing memories are associated with different phases of the theta rhythm in the human hippocampus. In brief, the authors use a proactive interference task in which subjects are asked to associate a word with two competing images and then subsequently recall the most recent image. Using pattern classifiers on the MEG data, the authors are able to decode reactivated content of the target and competitor memories and find that these patterns appear locked to different phases of the hippocampal rhythm. They also show that those subjects with worse memory performance had fewer differences in the phases to which target and competitor memories are locked. Together, the data provide support for a computational model of competing memories which suggests that oscillatory inhibition can be leveraged to strengthen or weaken inhibition of target and competitor memories (oscillating interference resolution model). One of the main strengths of the manuscript is that this is a pre-registered study, and so the specific hypotheses tested here have previously been reported. The current manuscript does not deviate too significantly from the pre-registered hypotheses and plan and reports the results of those proposed analyses. As such, this manuscript, therefore, presents a valuable addition to the literature, since it reports the results of a clearly established set of hypotheses testing a very specific question regarding memory interference.
-