Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Large-Scale Automated Sleep Staging.

Haoqi Sun1,2, Jian Jia3, Balaji Goparaju4

  • 1Energy Research Institute @ NTU, Interdisciplinary Graduate School, Nanyang Technological University, 639798, Singapore.

Sleep
|October 14, 2017
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

What Is the Effect of Robot Reduction in Displaced Pelvic Fractures? A Multicenter Randomized Clinical Trial.

Clinical orthopaedics and related research·2026
Same author

Automated Prediction of Glasgow Coma Scale Scores From Unstructured Electronic Health Records Using Natural Language Processing: Development and Validation Study.

Journal of medical Internet research·2026
Same author

Delayed Arousal Response to Sleep Apnea Encodes Mortality.

medRxiv : the preprint server for health sciences·2026
Same author

Delirium and Increased Risk of Developing Dementia: An Emulated Target Trial Analysis.

medRxiv : the preprint server for health sciences·2026
Same author

Artificial Intelligence Diagnosis of Obstructive Sleep Apnoea using Overnight Pulse Oximetry: A Systematic Review and Bayesian Meta-Analysis.

Journal of medical Internet research·2026
Same author

Complexity of resting cortical activity predicts neurophysiological responses to theta-burst stimulation but fails to generalize: A rigorous machine-learning approach.

PLoS computational biology·2026
Same journal

Nighttime light exposure is associated with metabolic dysfunction in schizophrenia: A cross-sectional analysis of the LENS study.

Sleep·2026
Same journal

Sleep Need Outcompetes Preparation: Reframing Sleep Initiation Through Naturalistic Behaviour.

Sleep·2026
Same journal

The Quest for Automated Pediatric Sleep Scoring: Are We There Yet?

Sleep·2026
Same journal

Sex Differences in the Sleep Architecture and Sleep-Disordered Breathing in C57BL/6 J Mice.

Sleep·2026
Same journal

Differential Effects of Prenatal Depression and Anxiety on Infant Sleep: Dual-Pathway Mechanisms Involving the HPA Axis and the Gut-Brain Axis.

Sleep·2026
Same journal

Metabolic Syndrome and Obstructive Sleep Apnea: Two Sides of the Same Coin.

Sleep·2026
See all related articles

Automated sleep staging using machine learning shows performance comparable to human scorers with sufficient training data. Increasing training data, contextual information, and model complexity improved accuracy, with results likely to generalize broadly.

Area of Science:

  • Computational neuroscience
  • Medical informatics
  • Sleep medicine

Background:

  • Automated sleep staging is challenged by clinical and physiological heterogeneity.
  • Large datasets can address these limitations through robust calibration.
  • The impact of sample size on automated staging performance remains unclear.

Purpose of the Study:

  • To assess machine learning performance in approximating human sleep staging with sufficient training data.
  • To investigate the influence of training set size, contextual information, model complexity, and data imbalance on staging accuracy.

Main Methods:

  • Extracted 102 features from six electroencephalography (EEG) channels.
  • Utilized 2000 nights of polysomnography data, split into 1000 for training and 1000 for testing.
Keywords:
EEGbig datamachine learningsleep stages

Related Experiment Videos

  • Measured agreement using epoch-by-epoch Cohen's kappa statistics against American Academy of Sleep Medicine criteria.

    • Main Results:

    • Cohen's kappa improved with training data up to approximately 300 recordings, then saturated.
    • Contextual information, increased model complexity, and adjusted training for imbalanced stages further enhanced performance.
    • The final model achieved a Cohen's kappa of 0.68 on the testing set, with lower variance for larger test sets.

    Conclusions:

    • Large-scale training data enables automated sleep staging that rivals human scorer performance.
    • The study's findings, validated on a large, heterogeneous dataset, suggest broad generalizability of the automated staging model.