skip to main content
Language:
Search Limited to: Search Limited to: Resource type Show Results with: Show Results with: Search type Index

0307 Automated Sleep Stage Scoring Using Deep Learning

Sleep (New York, N.Y.), 2018-04, Vol.41 (suppl_1), p.A118-A118 [Peer Reviewed Journal]

Sleep Research Society 2018. Published by Oxford University Press [on behalf of the Sleep Research Society]. All rights reserved. For permissions, please email: journals.permissions@oup.com 2018 ;Copyright © 2018 Sleep Research Society ;ISSN: 0161-8105 ;EISSN: 1550-9109 ;DOI: 10.1093/sleep/zsy061.306

Full text available

Citations Cited by
  • Title:
    0307 Automated Sleep Stage Scoring Using Deep Learning
  • Author: Zhang, L ; Fabbri, D ; Upender, R ; Kent, D
  • Subjects: Accuracy ; Automation ; Neural networks ; Sleep
  • Is Part Of: Sleep (New York, N.Y.), 2018-04, Vol.41 (suppl_1), p.A118-A118
  • Description: Abstract Introduction Polysomnography (PSG) scoring is labor intensive and suffers from considerable variability in inter- and intra-rater reliability. Automated PSG scoring has the potential to reduce the human labor costs and variability inherent to this task. Deep learning is a form of machine learning that uses neural networks to recognize data patterns by inspecting many examples rather than by following explicit programming. It uses layers of perceptrons (probabilistic classifiers) to learn hierarchical structures and dependencies within the data. Deep learning can be used to build highly accurate classifiers out of noisy, real-world datasets. Methods This study performed a retrospective evaluation of scored PSG data from the Sleep Heart Health Study (SHHS). A sleep staging classifier was developed with deep learning using a combination of convolutional and recurrent layers. The training set was composed of 1,900,000 30-second PSG epochs and their associated stage scores from randomly selected patients within the SHHS dataset. Spectrograms were generated from electroencephalography, electrooculography, and electromyography data using fast Fourier transforms over time for each epoch and then passed to the neural network. A hold-out set of 200,000 PSG epochs from randomly selected non-training patients was used to assess model accuracy and discrimination via F1-score, per-stage accuracy and Cohen’s Kappa (K). Results The neural network classifier demonstrated an F1-score of 0.86. Accuracy for the model based on the manual scores was calculated for the W (94.0%), N1 (40.0%), N2 (90.4%), N3 (72.6%), and R (86.5%) stages. The agreement between the model and manual scores was calculated as K = 0.8. Conclusion The deep learning sleep stage classifier demonstrates excellent accuracy and agreement with expert sleep stage scoring. It achieves the best overall F1-score, accuracy and Cohen’s Kappa compared to literature for automated sleep stage scoring of PSG epochs. Accurate automated scoring of other PSG events may eventually allow for fully automated PSG scoring. Support (If Any) None.
  • Publisher: US: Oxford University Press
  • Language: English
  • Identifier: ISSN: 0161-8105
    EISSN: 1550-9109
    DOI: 10.1093/sleep/zsy061.306
  • Source: ProQuest One Psychology
    Alma/SFX Local Collection
    ProQuest Central
Back to results list
INSPIRE LIBRARY - TON DUC THANG UNIVERSITY (84-028) 37 755 057 Feedback
19 Nguyen Huu Tho St. Dist.7, HCM thuvien@tdtu.edu.vn

Copyright © 2017 INSPIRE LIBRARY - TON DUC THANG UNIVERSITY

Searching Remote Databases, Please Wait