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

The DOI matches its Crossref record; 75 of the 75 references it cites were found in a registry or an index; 41 of the 63 sources its claims cite were checked for support.

Naturalistic Hyperscanning with Wearable Magnetoencephalography

Open access

Published version of a preprint

Abstract

The evolution of human cognitive function is reliant on complex social interactions which form the behavioural foundation of who we are. These social capacities are subject to dramatic change in disease and injury; yet their supporting neural substrates remain poorly understood. Hyperscanning employs functional neuroimaging to simultaneously assess brain activity in two individuals and offers the best means to understand the neural basis of social interaction. However, present technologies are limited, either by poor performance (low spatial/temporal precision) or an unnatural scanning environment (claustrophobic scanners, with interactions via video). Here, we describe hyperscanning using wearable magnetoencephalography (MEG) based on optically pumped magnetometers (OPMs). We demonstrate our approach by simultaneously measuring brain activity in two subjects undertaking two separate tasks-an interactive touching task and a ball game. Despite large and unpredictable subject motion, sensorimotor brain activity was delineated clearly, and the correlation of the envelope of neuronal oscillations between the two subjects was demonstrated. Our results show that unlike existing modalities, OPM-MEG combines high-fidelity data acquisition and a naturalistic setting and thus presents significant potential to investigate neural correlates of social interaction.

Overview

CoChat Summary

TL;DR

This paper reports the first hyperscanning experiments using wearable magnetoencephalography (MEG) based on optically pumped magnetometers (OPMs), with two participants scanned simultaneously during naturalistic tasks. It demonstrates that OPM-MEG can recover sensorimotor beta-band activity and inter-brain oscillatory relationships despite substantial participant motion.

Key findings

  • Matrix-coil field nulling reduced remnant fields over the helmets: in the touching task, from 3.3 ± 1.8 nT to 0.48 ± 0.44 nT (factor 6.9) for Participant 1 and from 4.1 ± 2.2 nT to 0.43 ± 0.40 nT (factor 9.5) for Participant 2; in the ball game, from 3.8 ± 2.8 nT to 1.6 ± 1.3 nT (factor 2.4) for Participant 1 and from 5.7 ± 2.9 nT to 2.0 ± 1.2 nT (factor 2.9) for Participant 3.
  • Despite motion, beamformer images showed sensorimotor/motor activation and task-induced beta-amplitude reduction. In the touching task, maximum translations were 16 mm and 24 mm and maximum rotations 3.0° and 7.9° for Participants 1 and 2; in the ball game, maximum translations were 50.0 mm and 64.8 mm and maximum rotations 17.0° and 17.4° for Participants 1 and 3.
  • In the touching task, the active participant showed beta-power reduction that began earlier and persisted longer than in the passive participant.
  • In the ball game, beta envelopes from the two participants evolved in anti-correlation with a lag of ~0.6 s between participants.

Methods

  • OPM-MEG system: up to 50 second-generation QuSpin zero-field magnetometers, with 16 OPMs per participant over left sensorimotor cortex in 3D-printed helmets; data sampled at 1200 Hz; participant motion tracked optically at 120 Hz. A 48-channel bi-planar matrix coil system (24 square coils per plane) was used for active field nulling.
  • Tasks: a two-person touching task with 30 repetitions (60 trials total) and 5 s inter-trial interval; and a two-person table-tennis ball game with 25 repetitions of 5 s play and 7 s rest, excluding two dropped-ball trials.
  • Analysis: linearly constrained minimum variance beamformer, pseudo-T-statistical contrast between active and control windows, time–frequency analysis of beta band (13–30 Hz), virtual electrode reconstruction, and autocorrelation/cross-correlation of beta envelopes.

Limitations

  • The paper states that participant movement away from the nulled volumes was deliberately limited, with participants’ feet kept planted, because large field changes from sensor movement through a non-zero field can cause OPM nonlinearities, cross-axis projection errors, sensor gain changes, and reduced neuronal source-modelling accuracy.
  • The paper states that extending the dynamic field-control approach to two helmets will be required to enable the full potential of OPM-MEG hyperscanning, and that the presented demonstrations only hint at the possibilities.
  • Evident from the text: the experiments used very small samples (two participants per task, with Participant 1 reused across tasks), so no group-level statistical inference is possible; ball-game nulling performance was lower than in the touching task because the larger participant separation meant fewer unit coils contributed to nulling.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3390/s23125454

Cites

75 of 75 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 A schematic representation of the OPM-MEG system. The system is housed in a magnetically shielded room (MSR). OPMs are interfaced with a series of data acquisition devices. Data from the OPMs are used to drive the matrix coil field nulling process, before an MEG recording begins. Optical tracking of the helmets is performed to monitor motion during a session. Instruction is passed to the participants via auditory cues controlled using a separate stimulus PC.
Figure 1 A schematic representation of the OPM-MEG system. The system is housed in a magnetically shielded room (MSR). OPMs are interfaced with a series of data acquisition devices. Data from the OPMs are used to drive the matrix coil field nulling process, before an MEG recording begins. Optical tracking of the helmets is performed to monitor motion during a session. Instruction is passed to the participants via auditory cues controlled using a separate stimulus PC.CC BY · PubMed Central
Figure 2 The position and orientation of the OPMs used in each experiment. Sensors that were used to inform the field nulling are shown in blue and additional sensors are shown in red. OPMs were concentrated over the left sensorimotor cortex, with additional sensors placed at the right and the front of the head, to inform the nulling process. (a) Sensor layout for Participant 1 during the two-person touch and ball game tasks. (b) Sensor layout for Participant 2 during the two-person touching task. (c) Sensor layout for Participant 3 during the two-person ball game.
Figure 2 The position and orientation of the OPMs used in each experiment. Sensors that were used to inform the field nulling are shown in blue and additional sensors are shown in red. OPMs were concentrated over the left sensorimotor cortex, with additional sensors placed at the right and the front of the head, to inform the nulling process. (a) Sensor layout for Participant 1 during the two-person touch and ball game tasks. (b) Sensor layout for Participant 2 during the two-person touching task. (c) Sensor layout for Participant 3 during the two-person ball game.CC BY · PubMed Central
Figure 3 Effect of field nulling. (a) The strength of the DC field, reported by the 48 total field measurements from 24 OPMs (12 per participant), with and without the matrix coils active, during the 2-person touch task. The error bars show standard errors across sensors. (b) Equivalent to (a) for the 2-person ball game.
Figure 3 Effect of field nulling. (a) The strength of the DC field, reported by the 48 total field measurements from 24 OPMs (12 per participant), with and without the matrix coils active, during the 2-person touch task. The error bars show standard errors across sensors. (b) Equivalent to (a) for the 2-person ball game.CC BY · PubMed Central
Figure 4 OPM-MEG data collected during a two-person naturalistic touching experiment. (a) Two participants stood on either side of a table. In odd-numbered trials, Participant 1 strokes the right hand of Participant 2, with their right hand. In even-numbered trials, the roles are reversed. (b) Beamformer images show the spatial signature of beta band modulation (thresholded to 80% of the maximum value). The odd trials (Participant 1 active) are shown in red and the even trials (Participant 2 active) are shown in blue. The spatial pattern suggests activity in the sensorimotor regions. (Note that there is a large overlap, so the blue overlay is partially obscured). (c) Timecourses showing the trial-averaged envelope of beta oscillations, extracted from the left sensorimotor cortex (peak in the beamformer images). Data from Participants 1 and 2 are shown in the top and bottom rows, respectively. In both cases, the red trace shows data recorded when Participant 1 was touching Participant 2. Similarly, (d,e) show time-frequency spectra of virtual electrodes at the peak of the beamformer images for each participant for both cases. Black dashed lines show the beta band.
Figure 4 OPM-MEG data collected during a two-person naturalistic touching experiment. (a) Two participants stood on either side of a table. In odd-numbered trials, Participant 1 strokes the right hand of Participant 2, with their right hand. In even-numbered trials, the roles are reversed. (b) Beamformer images show the spatial signature of beta band modulation (thresholded to 80% of the maximum value). The odd trials (Participant 1 active) are shown in red and the even trials (Participant 2 active) are shown in blue. The spatial pattern suggests activity in the sensorimotor regions. (Note that there is a large overlap, so the blue overlay is partially obscured). (c) Timecourses showing the trial-averaged envelope of beta oscillations, extracted from the left sensorimotor cortex (peak in the beamformer images). Data from Participants 1 and 2 are shown in the top and bottom rows, respectively. In both cases, the red trace shows data recorded when Participant 1 was touching Participant 2. Similarly, (d,e) show time-frequency spectra of virtual electrodes at the peak of the beamformer images for each participant for both cases. Black dashed lines show the beta band.CC BY · PubMed Central
Figure 5 OPM-MEG data collected during a two-person ball game. (a) Two participants each held a table-tennis bat in their right hand and hit the ball back and forth to each other; a 5 s rally was followed by 7 s rest. (b) Beamformer images of beta band modulation between task and rest (thresholded to 70% of the maximum value). The spatial pattern suggests beta power reduction in the sensorimotor regions during the rally. (c) Time–frequency spectrograms were extracted from the left sensorimotor cortex for Participant 1 (top) and Participant 3 (bottom). Black dashed lines show the beta band. (d) Timecourses of the envelope of beta band activity (again Participant 1 top, and Participant 3 bottom). Data suggest anti-correlation between 3 and 6.5 s (marked with black dashed lines). (e) Inset: the timecourses extracted in the 3–6.5 s window and overlaid. Main: comparison of the autocorrelations of the two extracted timecourses (blue/red) with their cross-correlation (yellow) reveals anti-correlation with a lag of ~0.6 s between the participant’s brain activity. The photographs shown are of the authors.
Figure 5 OPM-MEG data collected during a two-person ball game. (a) Two participants each held a table-tennis bat in their right hand and hit the ball back and forth to each other; a 5 s rally was followed by 7 s rest. (b) Beamformer images of beta band modulation between task and rest (thresholded to 70% of the maximum value). The spatial pattern suggests beta power reduction in the sensorimotor regions during the rally. (c) Time–frequency spectrograms were extracted from the left sensorimotor cortex for Participant 1 (top) and Participant 3 (bottom). Black dashed lines show the beta band. (d) Timecourses of the envelope of beta band activity (again Participant 1 top, and Participant 3 bottom). Data suggest anti-correlation between 3 and 6.5 s (marked with black dashed lines). (e) Inset: the timecourses extracted in the 3–6.5 s window and overlaid. Main: comparison of the autocorrelations of the two extracted timecourses (blue/red) with their cross-correlation (yellow) reveals anti-correlation with a lag of ~0.6 s between the participant’s brain activity. The photographs shown are of the authors.CC BY · PubMed Central

Verification

Claims the paper makes

24 Valid · 1 Not valid · 12 Not checked · 33 Own results and interpretations

37 more claims · 9 with their cited sources checked

Create a free account to see all 70 claims and to verify the 4 references still open.

See all claims — free account
  • It may state two separate claims as one. 15
  • Probably not a substantial claim. 17

Create a free account to see these claims.

Self-check

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

Each contradiction between two passages listed here was confirmed by an independent check.

1 of 1

Citations

Cites

Show all 59 references

Cited by

See all 52

Metadata

Authors
Holmes, N., Rea, M., Hill, R. M., Boto, E., Leggett, J., Edwards, L. J., Rhodes, N., Shah, V., Osborne, J., Fromhold, T. M., Glover, P., Montague, P. R., Brookes, M. J., & Bowtell, R.
Affiliations
  • University of Nottingham (GB)
  • QuSpin (United States) (US)
Corresponding author
Holmes, Niall
Venue
Sensors
Publisher
MDPI AG
Published
2023-06-09
Volume
23
Issue
12
Pages
5454
ISSN
1424-8220 (ISSN-L)
OpenAlex
W4380082698
Cited by
52
References
59
Citations by year
2026: 12 · 2025: 10 · 2024: 14 · 2023: 8
FWCI
5.21
Type
journal article
Language
en
Open access
Locations
  • Sensors (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Repository@Nottingham (University of Nottingham) (publishedVersion, other-oa)
  • VTechWorks (Virginia Tech) (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • National Institutes of Health (R01EB028772, EP/T001046/1, EP/V047264/1, 203257/Z/16/Z, 203257/B/16/Z)
  • the UK Quantum Technology Hub in Sensing and Timing (R01EB028772, EP/T001046/1, EP/V047264/1, 203257/Z/16/Z, 203257/B/16/Z)
  • Healthcare Impact Partnership (R01EB028772, EP/T001046/1, EP/V047264/1, 203257/Z/16/Z, 203257/B/16/Z)
  • the UK Engineering and Physical Sciences Research Council (EPSRC) (R01EB028772, EP/T001046/1, EP/V047264/1, 203257/Z/16/Z, 203257/B/16/Z)
  • a Wellcome Collaborative Award in Science (R01EB028772, EP/T001046/1, EP/V047264/1, 203257/Z/16/Z, 203257/B/16/Z)
Topics
  • Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Atomic and Subatomic Physics Research
  • Life Sciences › Neuroscience › Cognitive Neuroscience › EEG and Brain-Computer Interfaces
  • Social Sciences › Psychology › Social Psychology › Action Observation and Synchronization
Keywords
hyperscanning, optically pumped magnetometers, social interaction, subject motion, ball games
MeSH terms
Wearable Electronic Devices, Brain, Humans, Magnetoencephalography, Functional Neuroimaging

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.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol and qwen3.8-omni-flash).

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: Naturalistic Hyperscanning with Wearable Magnetoencephalography. https://app.cochat.ai/papers/10.3390/s23125454
The paper
Holmes, N., Rea, M., Hill, R. M., Boto, E., Leggett, J., Edwards, L. J., Rhodes, N., Shah, V., Osborne, J., Fromhold, T. M., Glover, P., Montague, P. R., Brookes, M. J., & Bowtell, R. (2023). Naturalistic Hyperscanning with Wearable Magnetoencephalography. Sensors. https://doi.org/10.3390/s23125454

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.