Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 49 of the 49 references it cites were found in a registry or an index.

Electrophysiological mechanisms of human memory consolidation

Open access
Zhang, Hui, Fell, Juergen, Axmacher, Nikolai · 2018 · Nature Communications

Abstract

Consolidation stabilizes memory traces after initial encoding. Rodent studies suggest that memory consolidation depends on replay of stimulus-specific activity patterns during fast hippocampal “ripple” oscillations. Here, we measured replay in intracranial electroencephalography recordings in human epilepsy patients, and related replay to ripples. Stimulus-specific activity was identified using representational similarity analysis and then tracked during waking rest and sleep after encoding. Stimulus-specific gamma (30–90 Hz) activity during early (100–500 ms) and late (500–1200 ms) encoding is spontaneously reactivated during waking state and sleep, independent of later memory. Ripples during nREM sleep, but not during waking state, trigger replay of activity from the late time window specifically for remembered items. Ripple-triggered replay of activity from the early time window during nREM sleep is enhanced for forgotten items. These results provide the first electrophysiological evidence for replay related to memory consolidation in humans, and point to a prominent role of nREM ripple-triggered replay in consolidation processes.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41467-018-06553-y

Cites

49 of 49 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Analysis of stimulus-specific representations. a Experimental design (day with nap). b Example of distributed activity in the gamma frequency range (30–90 Hz) during encoding and retrieval of the same item. The right panel shows the same example data as a scatter plot. Each sphere corresponds to one electrode contact. c Analysis of stimulus-specific representations: schematic overview. EEG power time series were extracted during encoding and retrieval of each item for each channel and each frequency band. EEG data were segmented into 200 ms windows with 100 ms overlap, and averaged within each window and each channel for both encoding and retrieval. This resulted in a 13 (window) × n (channel) × 2 (encoding, retrieval) matrix for each item and each patient. We then performed Spearman’s correlations across channels separately for each window during encoding and each window during retrieval of the same item. This resulted in one 13 (encoding window) × 13 (retrieval window) correlation matrix for each item and each patient. Fisher-Z transformed correlation matrices were averaged across items within each patient, resulting in one 13 (encoding window) × 13 (retrieval window) correlation matrix (RSAsame) for each patient. A similar analysis was conducted between encoding of one item and retrieval of a different item. This resulted in another 13 (encoding window) × 13 (retrieval window) correlation matrix (RSAdiffer) for each patient. We then performed paired t tests between RSAsame and RSAdiffer across patients. We used a cluster-based surrogate analysis for correction of multiple comparisons. Surrogate data were generated by randomly shuffling item labels and performing the same RSA procedure as described above. We extracted clusters from these surrogate matrices and selected for each surrogate the cluster with the largest summary t value. This surrogate procedure was repeated 10,000 times and produced 10,000 surrogate clusters. Then, we ranked each empirical cluster within the distribution of all surrogate clusters. rg indicates the ranking value. For details, see Methods
Fig. 1 Analysis of stimulus-specific representations. a Experimental design (day with nap). b Example of distributed activity in the gamma frequency range (30–90 Hz) during encoding and retrieval of the same item. The right panel shows the same example data as a scatter plot. Each sphere corresponds to one electrode contact. c Analysis of stimulus-specific representations: schematic overview. EEG power time series were extracted during encoding and retrieval of each item for each channel and each frequency band. EEG data were segmented into 200 ms windows with 100 ms overlap, and averaged within each window and each channel for both encoding and retrieval. This resulted in a 13 (window) × n (channel) × 2 (encoding, retrieval) matrix for each item and each patient. We then performed Spearman’s correlations across channels separately for each window during encoding and each window during retrieval of the same item. This resulted in one 13 (encoding window) × 13 (retrieval window) correlation matrix for each item and each patient. Fisher-Z transformed correlation matrices were averaged across items within each patient, resulting in one 13 (encoding window) × 13 (retrieval window) correlation matrix (RSAsame) for each patient. A similar analysis was conducted between encoding of one item and retrieval of a different item. This resulted in another 13 (encoding window) × 13 (retrieval window) correlation matrix (RSAdiffer) for each patient. We then performed paired t tests between RSAsame and RSAdiffer across patients. We used a cluster-based surrogate analysis for correction of multiple comparisons. Surrogate data were generated by randomly shuffling item labels and performing the same RSA procedure as described above. We extracted clusters from these surrogate matrices and selected for each surrogate the cluster with the largest summary t value. This surrogate procedure was repeated 10,000 times and produced 10,000 surrogate clusters. Then, we ranked each empirical cluster within the distribution of all surrogate clusters. rg indicates the ranking value. For details, see MethodsCC BY · PubMed Central
Fig. 2 Identification of stimulus-specific activity patterns. a Two clusters showed stimulus-specific representations (i.e., higher correlations between encoding of one and retrieval of the same item as compared to encoding of one and retrieval of a different item) in the gamma range: an early cluster (100–500 ms, blue frame) and a late cluster (500–1200 ms, black frame). b Functional relevance of clusters showing stimulus-specific representations in panel a for memory. Higher similarity between encoding and retrieval of subsequently remembered than forgotten items in the late cluster, but not in the early cluster. Each colorful dot indicates one participant. Same colors indicate data from the same participant. c Two clusters showed stimulus-specific representations in the epsilon range. d The similarity level within clusters in c did not differ between remembered and forgotten items. Each colorful dot indicates one participant. Same colors indicate data from the same participant. Error bars, standard error of the mean; **p < 0.01 (paired t test)
Fig. 2 Identification of stimulus-specific activity patterns. a Two clusters showed stimulus-specific representations (i.e., higher correlations between encoding of one and retrieval of the same item as compared to encoding of one and retrieval of a different item) in the gamma range: an early cluster (100–500 ms, blue frame) and a late cluster (500–1200 ms, black frame). b Functional relevance of clusters showing stimulus-specific representations in panel a for memory. Higher similarity between encoding and retrieval of subsequently remembered than forgotten items in the late cluster, but not in the early cluster. Each colorful dot indicates one participant. Same colors indicate data from the same participant. c Two clusters showed stimulus-specific representations in the epsilon range. d The similarity level within clusters in c did not differ between remembered and forgotten items. Each colorful dot indicates one participant. Same colors indicate data from the same participant. Error bars, standard error of the mean; **p < 0.01 (paired t test)CC BY · PubMed Central
Fig. 3 Spontaneous replay during rest. a Schematic illustration of analysis procedure. Reactivation is quantified by correlating activity during each encoding event and each time period during the resting state. Each circle in the brain indicates one channel. The hotter the color of the circle, the higher the EEG power. b Higher replay of stimulus-specific activity from late encoding cluster (500–1200 ms) of remote vs. recent items. * indicates p < 0.05 (paired t test). Reactivation does not differ between remembered and forgotten remote items c or between waking state and nREM sleep d. e Encoding-time resolved replay during rest. Reactivation differs between remote and recent items in three time periods marked with *s in different colors after cluster-based correction for multiple comparisons (surrogate data was generated by switching remote and recent labels). f Averaged replay levels of all clusters depicted in e as compared to zero. Each colorful dot indicates one participant. Same colors indicate data from the same participant. Error bars, standard error of the mean; * indicates significant difference from zero (one sample t-test, p < 0.05)
Fig. 3 Spontaneous replay during rest. a Schematic illustration of analysis procedure. Reactivation is quantified by correlating activity during each encoding event and each time period during the resting state. Each circle in the brain indicates one channel. The hotter the color of the circle, the higher the EEG power. b Higher replay of stimulus-specific activity from late encoding cluster (500–1200 ms) of remote vs. recent items. * indicates p < 0.05 (paired t test). Reactivation does not differ between remembered and forgotten remote items c or between waking state and nREM sleep d. e Encoding-time resolved replay during rest. Reactivation differs between remote and recent items in three time periods marked with *s in different colors after cluster-based correction for multiple comparisons (surrogate data was generated by switching remote and recent labels). f Averaged replay levels of all clusters depicted in e as compared to zero. Each colorful dot indicates one participant. Same colors indicate data from the same participant. Error bars, standard error of the mean; * indicates significant difference from zero (one sample t-test, p < 0.05)CC BY · PubMed Central
Fig. 4 Ripple-triggered replay. a Grand average of ripples and ripple-triggered raw EEG traces in the hippocampus. Time 0 indicates the onset of ripple events. b Examples of individual ripple events in four patients (highpass filtered at 5 Hz for display purposes). The blue line marks the duration of ripple events. c Analysis of replay during ripple events and peri-ripple periods: schematic overview. Top: Example of intracranial EEG activity from the hippocampus (highpass filtered at 40 Hz). Below: time windows used to extract activity during peri-ripple periods and ripple events. d Time resolved replay of gamma activity locked to ripples during nREM sleep and waking state for later remembered and forgotten remote items. Ripple time 0 corresponds to ripple events and ripple time before and after 0 correspond to peri-ripple period. The early and the late cluster correspond to the stimulus-specific clusters in Fig. 2a. e Ripple-triggered replay of early and late encoding activity differentially affects memory: Interaction between replay levels of remembered and forgotten items and gamma band activity from the early (100–500 ms) and the late (500–1200 ms) encoding clusters. f Direct comparisons of replay levels during ripple events between later remembered and later forgotten remote items. Each colorful dot indicates one participant. Same colors indicate data from the same participant. Error bars, standard error of the mean; *p < 0.05 (paired t test); **p < 0.01 (2 × 2 repeated measures ANOVA); X, p = 0.071 (paired t test); time windows showing higher replay levels for remembered vs. forgotten items during ripple events are indicated by gray background
Fig. 4 Ripple-triggered replay. a Grand average of ripples and ripple-triggered raw EEG traces in the hippocampus. Time 0 indicates the onset of ripple events. b Examples of individual ripple events in four patients (highpass filtered at 5 Hz for display purposes). The blue line marks the duration of ripple events. c Analysis of replay during ripple events and peri-ripple periods: schematic overview. Top: Example of intracranial EEG activity from the hippocampus (highpass filtered at 40 Hz). Below: time windows used to extract activity during peri-ripple periods and ripple events. d Time resolved replay of gamma activity locked to ripples during nREM sleep and waking state for later remembered and forgotten remote items. Ripple time 0 corresponds to ripple events and ripple time before and after 0 correspond to peri-ripple period. The early and the late cluster correspond to the stimulus-specific clusters in Fig. 2a. e Ripple-triggered replay of early and late encoding activity differentially affects memory: Interaction between replay levels of remembered and forgotten items and gamma band activity from the early (100–500 ms) and the late (500–1200 ms) encoding clusters. f Direct comparisons of replay levels during ripple events between later remembered and later forgotten remote items. Each colorful dot indicates one participant. Same colors indicate data from the same participant. Error bars, standard error of the mean; *p < 0.05 (paired t test); **p < 0.01 (2 × 2 repeated measures ANOVA); X, p = 0.071 (paired t test); time windows showing higher replay levels for remembered vs. forgotten items during ripple events are indicated by gray backgroundCC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 49 references

Cited by

See all 171

Metadata

Authors
Zhang, H., Fell, J., & Axmacher, N.
Affiliations
  • Ruhr University Bochum (DE)
  • University of Bonn (DE)
Corresponding author
  • Zhang, Hui
  • Axmacher, Nikolai
Venue
Nature Communications
Published
2018-10-05
Volume
9
Issue
1
Pages
4103-4103
ISSN
2041-1723 (ISSN-L)
OpenAlex
W2894915550
Cited by
164
References
49
Citations by year
2026: 15 · 2025: 22 · 2024: 25 · 2023: 27 · 2022: 21 · 2021: 17 · 2020: 21 · 2019: 13 · 2018: 3
FWCI
8.04
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, other-oa)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • Universitätsbibliographie, Universität Duisburg-Essen (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
Deutsche Forschungsgemeinschaft (FE366/9-1, SFB1089, SFB 1280, SFB 874, AX82/3, SFB 1280/A02)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Memory and Neural Mechanisms
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Sleep and Wakefulness Research
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Memory Processes and Influences
Keywords
memory consolidation, neural replay, hippocampal ripples, NREM sleep, gamma oscillations, representational similarity analysis, intracranial EEG, forgetting, hippocampus
MeSH terms
Memory Consolidation, Adult, Electroencephalography, Female, Hippocampus, Humans, Male, Middle Aged, Sleep, REM

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Electrophysiological mechanisms of human memory consolidation. https://app.cochat.ai/papers/10.1038/s41467-018-06553-y
The paper
Zhang, H., Fell, J., & Axmacher, N. (2018). Electrophysiological mechanisms of human memory consolidation. Nature Communications. https://doi.org/10.1038/s41467-018-06553-y

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.