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

The DOI matches its Crossref record; 201 of the 202 references it cites were found in a registry or an index; 30 of the 88 sources its claims cite were checked for support.

The future of sleep health: a data-driven revolution in sleep science and medicine

Open access
Perez-Pozuelo, Ignacio, Zhai, Bing, Palotti, Joao, Mall, Raghvendra, Aupetit, Michaël, Garcia-Gomez, Juan M., Taheri, Shahrad, Guan, Yu, Fernandez-Luque, Luis · 2020 · npj Digital Medicine

Abstract

In recent years, there has been a significant expansion in the development and use of multi-modal sensors and technologies to monitor physical activity, sleep and circadian rhythms. These developments make accurate sleep monitoring at scale a possibility for the first time. Vast amounts of multi-sensor data are being generated with potential applications ranging from large-scale epidemiological research linking sleep patterns to disease, to wellness applications, including the sleep coaching of individuals with chronic conditions. However, in order to realise the full potential of these technologies for individuals, medicine and research, several significant challenges must be overcome. There are important outstanding questions regarding performance evaluation, as well as data storage, curation, processing, integration, modelling and interpretation. Here, we leverage expertise across neuroscience, clinical medicine, bioengineering, electrical engineering, epidemiology, computer science, mHealth and human-computer interaction to discuss the digitisation of sleep from a inter-disciplinary perspective. We introduce the state-of-the-art in sleep-monitoring technologies, and discuss the opportunities and challenges from data acquisition to the eventual application of insights in clinical and consumer settings. Further, we explore the strengths and limitations of current and emerging sensing methods with a particular focus on novel data-driven technologies, such as Artificial Intelligence.

Overview

CoChat Summary

## TL;DR

This is a narrative review that proposes a five-step “Digital Sleep Framework” for moving from sleep data acquisition to clinical and consumer applications. It argues that digital sleep technologies and AI could enable scalable, unobtrusive sleep monitoring, but only if major issues in validation, interoperability, bias, privacy, and reproducibility are addressed.

## Key findings

  • The paper frames sleep digitization around a five-step framework: data acquisition, data storage and curation, data pre-processing, AI-based sleep modelling, and data-driven applications.
  • PSG remains the gold standard, but it is “expensive, time-consuming” and impractical for long-term home use; ambulatory PSG is described as only a partial solution.
  • Actigraphy and accelerometry are recognized by the American Academy of Sleep Medicine as valid for sleep assessment, but the paper states three key limitations: lack of validation across consumer devices, lack of standardization in human-activity recognition, and lack of assessment techniques for daytime sleeping.
  • Some methods reported specific performance figures: a single-channel EEG ensemble model reached an agreement rate of 0.87 versus expert scoring, in-ear sensor sleep-stage prediction reached 74% agreement with full PSG, ultrasound-based movement detection showed 86% recall and <10% error, and iSleep reported >90% accuracy for event classification.

## Methods

  • Narrative review / perspective article synthesizing prior literature on sleep sensing, storage, preprocessing, modelling, and applications; no original participant study is reported.
  • The authors organize the discussion using a five-step “Digital Sleep Framework” spanning multi-modal acquisition through end-user application.
  • The analysis is qualitative and literature-based, citing examples across PSG, actigraphy, EEG, mobile sensing, bed sensors, WiFi/radio sensing, cloud/fog architectures, and machine learning/deep learning methods.

## Limitations

  • The paper explicitly notes that “few methods have been validated against gold-standard PSG or undergone systematic reliability assessment.”
  • It also states that consumer-oriented sleep technologies “lack validation” and that their underlying models “change frequently.”
  • Evident from the text: this is not an empirical evaluation of specific devices or algorithms; it is a broad synthesis, so its conclusions depend on the cited literature rather than new data collected in the paper.

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

Integrity signals

Registry

Crossref · 10.1038/s41746-020-0244-4

Cites

201 of 201 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1 The digital sleep framework covers the path of sleep data from its acquisition to when its insights are used for medical or consumer applications. The framework begins with the acquisition of sleep-related data. This can be done using a variety of sensors, ranging from polysomnography to bed sensors. This data is then stored and curated, a step that comprises privacy-aware storage, cleaning, filtering and anonymisation. Once that data has been appropriately treated, the processing step takes place whereby data is transformed and integrated based on the end-model. For example it may undergo different transformations like normalization or featurization. The next step entails modeling, which can consist of simple heuristic methods, statistical learning or deep learning methods, for example. Finally, the resulting model can be deployed for a variety either medical or consumer applications.
Fig. 1 The digital sleep framework covers the path of sleep data from its acquisition to when its insights are used for medical or consumer applications. The framework begins with the acquisition of sleep-related data. This can be done using a variety of sensors, ranging from polysomnography to bed sensors. This data is then stored and curated, a step that comprises privacy-aware storage, cleaning, filtering and anonymisation. Once that data has been appropriately treated, the processing step takes place whereby data is transformed and integrated based on the end-model. For example it may undergo different transformations like normalization or featurization. The next step entails modeling, which can consist of simple heuristic methods, statistical learning or deep learning methods, for example. Finally, the resulting model can be deployed for a variety either medical or consumer applications.CC BY · PubMed Central
Fig. 2 Emerging sleep-sensing technologies. Emerging sleep technologies range from non-contact methods like RF sensors to miniaturized, wireless or in-ear EEGs.
Fig. 2 Emerging sleep-sensing technologies. Emerging sleep technologies range from non-contact methods like RF sensors to miniaturized, wireless or in-ear EEGs.CC BY · PubMed Central
Fig. 3 Selected methods for the measurement of sleep and their accuracy and usability trade-off. This chart plots the accuracy of sleep-sensing methods at infering sleep-related metrics against their ease of use. For example, while polysomnography is considered the “gold-standard” technique to measure sleep, it is cumbersome and expensive.
Fig. 3 Selected methods for the measurement of sleep and their accuracy and usability trade-off. This chart plots the accuracy of sleep-sensing methods at infering sleep-related metrics against their ease of use. For example, while polysomnography is considered the “gold-standard” technique to measure sleep, it is cumbersome and expensive.CC BY · PubMed Central
Fig. 4 Holistic evaluation of sleep-monitoring methods. Some methods, such as PSG, are accurate but inappropriate for use in daily sleep monitoring, as they require professional set up and are intrusive. Other methods, such as bed sensors, are unobtrusive but more prone to noise than PSG.
Fig. 4 Holistic evaluation of sleep-monitoring methods. Some methods, such as PSG, are accurate but inappropriate for use in daily sleep monitoring, as they require professional set up and are intrusive. Other methods, such as bed sensors, are unobtrusive but more prone to noise than PSG.CC BY · PubMed Central
Fig. 5 Overview of cloud computing-based sleep data acquisition and storage. This illustration provides an overview of the process starting with device layer (which includes fast, real-time processing and data visualisation, embedded systems, gateways and micro data storage), followed by the fog layer (which includes local networks, virtualisation, data analysis and reduction) and finally cloud layer (which consists of data centres and big data storage and processing).
Fig. 5 Overview of cloud computing-based sleep data acquisition and storage. This illustration provides an overview of the process starting with device layer (which includes fast, real-time processing and data visualisation, embedded systems, gateways and micro data storage), followed by the fog layer (which includes local networks, virtualisation, data analysis and reduction) and finally cloud layer (which consists of data centres and big data storage and processing).CC BY · PubMed Central
Fig. 6 Sleep classification algorithms can be based on heuristic approaches or Artificial Intelligence. We describe machine learning/statistical learning approaches and deep-learning approaches within AI.
Fig. 6 Sleep classification algorithms can be based on heuristic approaches or Artificial Intelligence. We describe machine learning/statistical learning approaches and deep-learning approaches within AI.CC BY · PubMed Central
Fig. 7 Key areas of impact for sleep health. Emerging sleep health technologies will have an impact on patient monitoring, clinical care, insurance, the pharmaceutical industry and health and wellness applications, as well as other impacts including on digital therapeutics and sports performance.
Fig. 7 Key areas of impact for sleep health. Emerging sleep health technologies will have an impact on patient monitoring, clinical care, insurance, the pharmaceutical industry and health and wellness applications, as well as other impacts including on digital therapeutics and sports performance.CC BY · PubMed Central

Verification

Claims the paper makes

24 Valid · 16 Not valid · 41 Not checked · 47 Own results and interpretations

50 more claims · 12 with their cited sources checked

Create a free account to see all 128 claims and to verify the 14 references still open.

See all claims — free account
  • It may state two separate claims as one. 34
  • Probably not a substantial claim. 54

Create a free account to see these claims.

Citations

Cites

Show all 202 references

Research works printed but not linked: 4

  1. Zhao, M., Yue, S., Katabi, D., Jaakkola, T. S. & Bianchi, M. T. Learning sleep stages from radio signals: a conditional adversarial architecture. in Proceedings of the 34th International Conference on Machine Learning, 4100–4109 (2017). https://www.acm.org/publications.
  2. Van Drongelen, W. Signal Processing for Neuroscientists (Academic Press, 2018).
  3. Choi, Y. K. et al. Smartphone applications to support sleep self-management: review and evaluation. J. Clin. Sleep Med. 14, 1783–1790 (2018). · DOI
  4. IEEE. IEEE VIS 2015 Workshop on Personal Visualization: Exploring Data in Everyday Life. https://www.vis4me.com/personalvis15/papers.html (2015).

Cited by

See all 311

Metadata

Authors
Perez-Pozuelo, I., Zhai, B., Palotti, J., Mall, R., Aupetit, M., Garcia-Gomez, J. M., Taheri, S., Guan, Y., & Fernandez-Luque, L.
Affiliations
  • University of Cambridge (GB)
  • Newcastle University (GB)
  • Universitat Politècnica de València (ES)
  • Weill Cornell Medical College in Qatar (QA)
Corresponding author
  • Perez-Pozuelo, Ignacio
  • Mall, Raghvendra
Venue
npj Digital Medicine
Published
2020-03-23
Volume
3
Issue
1
Pages
42-42
ISSN
2398-6352 (ISSN-L)
OpenAlex
W3012581557
Cited by
311
References
202
Citations by year
2026: 35 · 2025: 71 · 2024: 55 · 2023: 51 · 2022: 48 · 2021: 38 · 2020: 11
FWCI
5.40
Type
journal article
Language
en
Open access
Licence
Locations
  • npj Digital Medicine (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Figshare (submittedVersion, cc-by)
  • Apollo (University of Cambridge) (submittedVersion, cc-by)
  • Apollo (cc-by)
  • Qatar National Library (cc-by)
Funders
RCUK | Engineering and Physical Sciences Research Council (iCase 17100053)
Topics
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Sleep and related disorders
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
  • Physical Sciences › Engineering › Biomedical Engineering › Non-Invasive Vital Sign Monitoring
Keywords
sleep monitoring, sleep health, wearable sensors, multi-sensor data, circadian rhythm, physical activity monitoring, sleep coaching, chronic disease, sleep epidemiology, mobile health, consumer health 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: The future of sleep health: a data-driven revolution in sleep science and medicine. https://app.cochat.ai/papers/10.1038/s41746-020-0244-4
The paper
Perez-Pozuelo, I., Zhai, B., Palotti, J., Mall, R., Aupetit, M., Garcia-Gomez, J. M., Taheri, S., Guan, Y., & Fernandez-Luque, L. (2020). The future of sleep health: a data-driven revolution in sleep science and medicine. npj Digital Medicine. https://doi.org/10.1038/s41746-020-0244-4

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.