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

The DOI matches its Crossref record; 38 of the 38 references it cites were found in a registry or an index; 14 of the 49 sources its claims cite were checked for support.

Miniaturized wireless, skin-integrated sensor networks for quantifying full-body movement behaviors and vital signs in infants

Open access

Abstract

Significance Early detection of infant neuromotor pathologies is critical for timely therapeutic interventions that rely on early-life neuroplasticity. Traditional assessments rely on subjective expert evaluations or specialized medical facilities, making them challenging to scale in remote and/or resource-constrained settings. The results presented here aim to democratize these evaluations using wireless networks of miniaturized, skin-integrated sensors that digitize movement behaviors and vital signs of infants in a cost-effective manner. The resulting data yield full-body motion reconstructions in the form of deidentified infant avatars, along with a range of important cardiopulmonary information. This technology approach enables rapid, routine evaluations of infants at any age via an engineering platform that has potential for use in nearly any setting across developed and developing countries alike.

Overview

CoChat Summary

TL;DR

The paper presents miniaturized, wireless, skin-integrated CORB sensors for quantifying full-body movement behaviors and vital signs in infants. It demonstrates feasibility in low-risk and elevated-risk infants across 1 wk, 1 mo, and 3 mo, with 3D motion reconstruction and cardiopulmonary monitoring.

Key findings

  • CORB sensors measure 32 × 21 × 3 mm, weigh less than 2.6 g, and are described as more than three times thinner, five times lighter, and two times smaller in volume than the Xsens MTw Awinda commercial motion-capture sensors.
  • Time-synchronized sensor data achieved submillisecond accuracy, with a representative four-sensor arm test showing less than 1 ms timing difference; reconstructed motion drift was 0.013°/s, and differences from video-based reconstruction increased from less than 0.2° initially to 1.3° over the measurement.
  • In representative 3-mo-old infants, quantitative activity level (QAL) was 14%, 66%, and 68% higher in the low-risk infant than the elevated-risk infant at 1 wk, 1 mo, and 3 mo, respectively; at 1 mo, head/chest QAL was 154% greater, arms 40% higher, and legs 64% higher in the low-risk infant.
  • Chest sensors captured heart rates of 120 to 160 beats per minute and respiration rates of 38 to 45 RPM; compared with an FDA-approved capnography device in three adults and 1,421 data points, respiratory rate showed a mean difference of 0.01 RPM and SD of 0.53 RPM.

Methods

  • The platform uses flexible printed circuit boards with a three-axis accelerometer and three-axis gyroscope, Bluetooth Low Energy, NAND flash memory, and a 7-mAh battery supporting more than 5 h of continuous operation; devices are encapsulated in silicone elastomer and attached with medical adhesive, bands, or wraps.
  • Human studies included two representative infants, one low-risk and one elevated-risk, assessed at approximately 1 wk, 1 mo, and 3 mo; 10 sensors were used at 3 mo and 6 sensors at earlier ages, with additional demonstrations in a 7-y-old child and adult respiratory-rate validation.
  • Motion reconstruction used synchronized inertial data, bias correction, low-pass filtering above 20 Hz, orientation estimation, and ROS/Rviz 3D avatar reconstruction; QAL was computed from spectral power between 0.1 and 10 Hz.

Limitations

  • The paper states that the infant results are feasibility demonstrations, that observed differences may reflect intersubject movement variability rather than underlying neuromotor disorders, and that larger-scale data are needed to define normative ranges.
  • The paper states that comparison respiratory data were not collected from infants because of size constraints with the capnography device used.
  • Evident from the text: the main QAL comparisons rely on two representative infants, motion reconstruction requires knowledge of initial body position and limb lengths, and the 200-Hz sampling setting prevents collection of phonocardiogram signals such as heart murmurs.

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

Integrity signals

Registry

Crossref · 10.1073/pnas.2104925118

Cites

38 of 38 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1. Images, schematic illustrations, and functional flow charts for miniaturized wireless sensors (i.e., CORB sensors) designed for quantifying gross motor behaviors and vital signs in infants. (A) Exploded-view schematic illustration of the device. Optical image of the device (Inset). (Scale bar, 1 cm.) (B) Measurement configuration for capturing full body motions, illustrated with a baby doll. (C–F) Images and finite element analysis computations of a device flexibility during various mechanical deformations: peeling (C and E) and bending with an angle of 90° (D and F). (G) Functional diagram of the platform showing hardware blocks including the power management, Bluetooth radio, microcontroller, memory, and six-axis inertial measurement unit for each device. The collected data from each device include synchronized timestamps to ensure millisecond relative timing accuracy. A user interface on a smartphone or table controls the devices, captures real-time data, and supports data downloads a 3D motion reconstruction using a local PC.
Fig. 1. Images, schematic illustrations, and functional flow charts for miniaturized wireless sensors (i.e., CORB sensors) designed for quantifying gross motor behaviors and vital signs in infants. (A) Exploded-view schematic illustration of the device. Optical image of the device (Inset). (Scale bar, 1 cm.) (B) Measurement configuration for capturing full body motions, illustrated with a baby doll. (C–F) Images and finite element analysis computations of a device flexibility during various mechanical deformations: peeling (C and E) and bending with an angle of 90° (D and F). (G) Functional diagram of the platform showing hardware blocks including the power management, Bluetooth radio, microcontroller, memory, and six-axis inertial measurement unit for each device. The collected data from each device include synchronized timestamps to ensure millisecond relative timing accuracy. A user interface on a smartphone or table controls the devices, captures real-time data, and supports data downloads a 3D motion reconstruction using a local PC.CC BY · PubMed Central
Fig. 2. Schematic illustrations and optical images of the sensor configuration and data flows for 3D motion reconstruction. (A) Schematic illustration of the flexible PCB that supports the power management and battery protection SoC, 4Gb NAND flash memory, six-axis inertial measurement unit, Bluetooth low energy SoC, and lithium-polymer battery. (B) Optical image of the fPCB after mounting all components (Left) and after folding (Right). (Scale bar, 1 cm.) (C) Image of the full system, including multiple time-synchronized devices, a charger, and user interface on a smartphone. (D) Time-synchronized and normalized accelerometer signals acquired from four sensors on the left upper arm (LUA), left lower arm (LLA), right upper arm (RUA), and right lower arm (RLA). The red arrow indicates the timing of a jumping movement. The results indicate time synchronized operation to within 10 ms or less. (E) Algorithm flowchart for motion reconstruction. The Top frames highlight the steps in bias correction, signal filtering, and time synchronization for each device. The Middle frames show static attitude extraction from the three-axis accelerometer and dynamic attitude extraction from the three-axis gyroscope. The Bottom frame corresponds to transfer of resulting data to the ROS for 3D motion reconstruction.
Fig. 2. Schematic illustrations and optical images of the sensor configuration and data flows for 3D motion reconstruction. (A) Schematic illustration of the flexible PCB that supports the power management and battery protection SoC, 4Gb NAND flash memory, six-axis inertial measurement unit, Bluetooth low energy SoC, and lithium-polymer battery. (B) Optical image of the fPCB after mounting all components (Left) and after folding (Right). (Scale bar, 1 cm.) (C) Image of the full system, including multiple time-synchronized devices, a charger, and user interface on a smartphone. (D) Time-synchronized and normalized accelerometer signals acquired from four sensors on the left upper arm (LUA), left lower arm (LLA), right upper arm (RUA), and right lower arm (RLA). The red arrow indicates the timing of a jumping movement. The results indicate time synchronized operation to within 10 ms or less. (E) Algorithm flowchart for motion reconstruction. The Top frames highlight the steps in bias correction, signal filtering, and time synchronization for each device. The Middle frames show static attitude extraction from the three-axis accelerometer and dynamic attitude extraction from the three-axis gyroscope. The Bottom frame corresponds to transfer of resulting data to the ROS for 3D motion reconstruction.CC BY · PubMed Central
Fig. 3. Representative data collected from a child subject during various arm and leg motions with different angles and orientations. (A) Normalized acceleration and gyroscope data during motion. The subject jumped (around 10 s), raised left and right arms to the front at 90° (around 19 s), and to the side at 135° (around 29 s) and 90° (around 39 s). (B) Optical and 3D reconstructed images from these motion data. 1: raised left and right arms to the front at 90°; 2: raised left and right arms to the side at 135°; 3: raised left and right arms to the side at 90°; 4: raised left arms to the side at 30°and right leg to the front at 90°; 5: raised left and right arms to the side at 45°and 120 °, respectively, and right leg to the side at 80°; and 6: raised left and right arms to the side at 90°and left leg to the side at 30°.
Fig. 3. Representative data collected from a child subject during various arm and leg motions with different angles and orientations. (A) Normalized acceleration and gyroscope data during motion. The subject jumped (around 10 s), raised left and right arms to the front at 90° (around 19 s), and to the side at 135° (around 29 s) and 90° (around 39 s). (B) Optical and 3D reconstructed images from these motion data. 1: raised left and right arms to the front at 90°; 2: raised left and right arms to the side at 135°; 3: raised left and right arms to the side at 90°; 4: raised left arms to the side at 30°and right leg to the front at 90°; 5: raised left and right arms to the side at 45°and 120 °, respectively, and right leg to the side at 80°; and 6: raised left and right arms to the side at 90°and left leg to the side at 30°.CC BY · PubMed Central
Fig. 4. Comparisons of representative gross motor behaviors associated with large physical motions from 3-mo-old infants at LR and ER of atypical motor development. LR and ER infants’ motor behaviors and corresponding optical images captured over the course of the 35-min session and associated 3D motion reconstruction results with different viewpoints. Captured time points for pictures and reconstruction results are indicated at the Bottom of each picture.
Fig. 4. Comparisons of representative gross motor behaviors associated with large physical motions from 3-mo-old infants at LR and ER of atypical motor development. LR and ER infants’ motor behaviors and corresponding optical images captured over the course of the 35-min session and associated 3D motion reconstruction results with different viewpoints. Captured time points for pictures and reconstruction results are indicated at the Bottom of each picture.CC BY · PubMed Central
Fig. 5. Quantitative comparisons of long-term and follow-up measurement results over a 3-mo period from representative infants at LR and ER of atypical motor development. (A) QAL comparisons of 3-mo-old LR and ER subjects in various postures such as prone, supine, supported sitting, supported standing, and horizontal suspension corresponding to the images in Fig 4. (B) 3D scatter plots from 3-mo-old LR and ER subjects‘ three-axis inertial measurements in supine position (magenta sphere). Acceleration of right and left arms from LR (first and second columns on the Top row) and ER subjects (first and second columns on the Bottom row). Angular velocity of right and left arms from LR (third and fourth columns on the Top row) and ER subjects (first and second columns on the Bottom row). XY (red), YZ (blue), and ZX (green) projections on each plot. (C) Overall QAL (Left) and each level from head and chest/arms/legs (Right) from the LR and ER infants when subjects are 1 wk, 1 mo, and 3 mo old.
Fig. 5. Quantitative comparisons of long-term and follow-up measurement results over a 3-mo period from representative infants at LR and ER of atypical motor development. (A) QAL comparisons of 3-mo-old LR and ER subjects in various postures such as prone, supine, supported sitting, supported standing, and horizontal suspension corresponding to the images in Fig 4. (B) 3D scatter plots from 3-mo-old LR and ER subjects‘ three-axis inertial measurements in supine position (magenta sphere). Acceleration of right and left arms from LR (first and second columns on the Top row) and ER subjects (first and second columns on the Bottom row). Angular velocity of right and left arms from LR (third and fourth columns on the Top row) and ER subjects (first and second columns on the Bottom row). XY (red), YZ (blue), and ZX (green) projections on each plot. (C) Overall QAL (Left) and each level from head and chest/arms/legs (Right) from the LR and ER infants when subjects are 1 wk, 1 mo, and 3 mo old.CC BY · PubMed Central
Fig. 6. Quantitative long-term and follow-up cardiac and respiratory activity measurement results over a 3-mo period from infants at LR and ER. (A) Normalized chest sensor signal from LR infants at 1 wk (Top), 1 mo (Middle), and 3 mo of age (Bottom). The black line corresponds to data from the z-axis of the accelerometer (seismocardiogram); the red line is the z-axis gyroscope signal; and the blue dashed line is a running average of this gyroscope signal (respiratory activity). AO and MO denote aortic and mitral valve opening, respectively. (B) Heart rate (Left) and respiratory rate (Right) determined from these data, LR (gray filling bar) and ER (red filling bar) subjects including error bars. (C) Time-synchronized full-body temperature monitoring from 3-mo-old LR infants collected by the CORB sensors. Temperature variation during data collection over 30 min (Left). Skin temperature determined from each device on the head, chest, and limbs including error bars (Right).
Fig. 6. Quantitative long-term and follow-up cardiac and respiratory activity measurement results over a 3-mo period from infants at LR and ER. (A) Normalized chest sensor signal from LR infants at 1 wk (Top), 1 mo (Middle), and 3 mo of age (Bottom). The black line corresponds to data from the z-axis of the accelerometer (seismocardiogram); the red line is the z-axis gyroscope signal; and the blue dashed line is a running average of this gyroscope signal (respiratory activity). AO and MO denote aortic and mitral valve opening, respectively. (B) Heart rate (Left) and respiratory rate (Right) determined from these data, LR (gray filling bar) and ER (red filling bar) subjects including error bars. (C) Time-synchronized full-body temperature monitoring from 3-mo-old LR infants collected by the CORB sensors. Temperature variation during data collection over 30 min (Left). Skin temperature determined from each device on the head, chest, and limbs including error bars (Right).CC BY · PubMed Central

Verification

Claims the paper makes

11 Valid · 3 Not valid · 14 Not checked · 65 Own results and interpretations

40 more claims · 3 with their cited sources checked

Create a free account to see all 93 claims and to verify the 2 references still open.

See all claims — free account
  • Probably not a substantial claim. 32
  • It may state two separate claims as one. 25

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.

3 of 3

Citations

Cites

Show all 38 references

Cited by

See all 59

Metadata

Authors
Jeong, H., Kwak, S. S., Sohn, S., Lee, J. Y., Lee, Y. J., O’Brien, M. K., Park, Y., Avila, R., Kim, J.-T., Yoo, J.-Y., Irie, M., Jang, H., Ouyang, W., Shawen, N., Kang, Y. J., Kim, S. S., Tzavelis, A., Lee, K., Andersen, R. A., Huang, Y., Jayaraman, A., Davis, M. M., Shanley, T., Wakschlag, L. S., Krogh-Jespersen, S., Xu, S., Ryan, S. W., Lieber, R. L., & Rogers, J. A.
Affiliations
  • Northwestern University (US)
  • Sibel (United States) (US)
  • University of Wisconsin–Madison (US)
  • Shirley Ryan AbilityLab (US)
  • Department of Medical Social Sciences, Feinberg School of Medicine, Northwestern University, Chicago, IL 60611;
Corresponding author
Rogers, John A.
Venue
Proceedings of the National Academy of Sciences
Published
2021-10-18
Volume
118
Issue
43
ISSN
  • 0027-8424 (ISSN-L)
  • 1091-6490
OpenAlex
W3206507112
Cited by
59
References
38
Citations by year
2026: 12 · 2025: 10 · 2024: 11 · 2023: 15 · 2022: 10 · 2021: 1
FWCI
3.32
Type
journal article
Language
en
Open access
Locations
  • Proceedings of the National Academy of Sciences (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • eScholarship (California Digital Library) (submittedVersion, cc-by)
Funders
Northwestern University
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
  • Physical Sciences › Engineering › Biomedical Engineering › Non-Invasive Vital Sign Monitoring
  • Physical Sciences › Engineering › Biomedical Engineering › Wireless Body Area Networks
Keywords
infant motor assessment, wireless inertial sensors, full-body kinematics, vital signs monitoring, gross motor skills, low-cost wearable sensors
MeSH terms
Child, Equipment Design, Heart Rate, Humans, Infant, Miniaturization, Monitoring, Physiologic, Movement, Skin, Video Recording, Bias, Infant Behavior, Imaging, Three-Dimensional, Vital Signs, Respiratory Rate, Wireless Technology

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).

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: Miniaturized wireless, skin-integrated sensor networks for quantifying full-body movement behaviors and vital signs in infants. https://app.cochat.ai/papers/10.1073/pnas.2104925118
The paper
Jeong, H., Kwak, S. S., Sohn, S., Lee, J. Y., Lee, Y. J., O’Brien, M. K., Park, Y., Avila, R., Kim, J.-T., Yoo, J.-Y., Irie, M., Jang, H., Ouyang, W., Shawen, N., Kang, Y. J., Kim, S. S., Tzavelis, A., Lee, K., Andersen, R. A., . . . Rogers, J. A. (2021). Miniaturized wireless, skin-integrated sensor networks for quantifying full-body movement behaviors and vital signs in infants. Proceedings of the National Academy of Sciences. https://doi.org/10.1073/pnas.2104925118

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.