Performance of an open machine learning model to classify sleep/wake from actigraphy across ∼24-hour intervals without knowledge of rest timing.

Autor: Roberts DM; Department of Biobehavioral Health, The Pennsylvania State University, University Park, Pennsylvania, USA; Proactive Life, Inc, New York, New York, USA. Electronic address: danmroberts@gmail.com., Schade MM; Department of Biobehavioral Health, The Pennsylvania State University, University Park, Pennsylvania, USA., Master L; Department of Biobehavioral Health, The Pennsylvania State University, University Park, Pennsylvania, USA., Honavar VG; Faculty of Data Sciences, College of Information Science and Technology, The Pennsylvania State University, University Park, Pennsylvania, USA., Nahmod NG; Department of Biobehavioral Health, The Pennsylvania State University, University Park, Pennsylvania, USA., Chang AM; Department of Biobehavioral Health, The Pennsylvania State University, University Park, Pennsylvania, USA., Gartenberg D; Proactive Life, Inc, New York, New York, USA., Buxton OM; Department of Biobehavioral Health, The Pennsylvania State University, University Park, Pennsylvania, USA.
Jazyk: angličtina
Zdroj: Sleep health [Sleep Health] 2023 Oct; Vol. 9 (5), pp. 596-610. Date of Electronic Publication: 2023 Aug 10.
DOI: 10.1016/j.sleh.2023.07.001
Abstrakt: Goal and Aims: Commonly used actigraphy algorithms are designed to operate within a known in-bed interval. However, in free-living scenarios this interval is often unknown. We trained and evaluated a sleep/wake classifier that operates on actigraphy over ∼24-hour intervals, without knowledge of in-bed timing.
Focus Technology: Actigraphy counts from ActiWatch Spectrum devices.
Reference Technology: Sleep staging derived from polysomnography, supplemented by observation of wakefulness outside of the staged interval. Classifications from the Oakley actigraphy algorithm were additionally used as performance reference.
Sample: Adults, sleeping in either a home or laboratory environment.
Design: Machine learning was used to train and evaluate a sleep/wake classifier in a supervised learning paradigm. The classifier is a temporal convolutional network, a form of deep neural network.
Core Analytics: Performance was evaluated across ∼24 hours, and additionally restricted to only in-bed intervals, both in terms of epoch-by-epoch performance, and the discrepancy of summary statistics within the intervals.
Additional Analytics and Exploratory Analyses: Performance of the trained model applied to the Multi-Ethnic Study of Atherosclerosis dataset.
Core Outcomes: Over ∼24 hours, the temporal convolutional network classifier produced the same or better performance as the Oakley classifier on all measures tested. When restricting analysis to the in-bed interval, the temporal convolutional network remained favorable on several metrics.
Important Supplemental Outcomes: Performance decreased on the Multi-Ethnic Study of Atherosclerosis dataset, especially when restricting analysis to the in-bed interval.
Core Conclusion: A classifier using data labeled over ∼24-hour intervals allows for the continuous classification of sleep/wake without knowledge of in-bed intervals. Further development should focus on improving generalization performance.
Competing Interests: Declaration of conflicts of interest DMR was employed by Proactive Life, Inc. at the time of the initial submission of the manuscript. At the point of revision, he was instead employed by Pennsylvania State University. DG is employed by Proactive Life, Inc., a for-profit company. Related to monitoring sleep, Proactive Life has two patents issued: Sleep Stimulation and Monitoring (US patent 10524661), Cyclical Behavior Modification (US Patent 8468115), four patents pending: Sleep Tracking Method and Device (16/950,987), Valence State Memory Association (16/504,285), Systems, Methods, and Apparatus for Monitoring Sleep (PCT/US21/59978), and two design patents pending: Holder for a Mobile Phone (29/796,124), Stands for Mobile Telephones (WIPO111054). Outside of the current work (last 4 years), Orfeu M. Buxton received honoraria/travel support for lectures/consulting from Boston University, Boston College, Tufts School of Dental Medicine, New York University, University of Miami, University of Utah, University of South Florida, University of Arion, Eric Angle Society for Orthodontists, and Allstate, consulting fees from SleepNumber, and receives an honorarium for his role as the Editor in Chief of Sleep Health (sleephealthjournal.org).
(Copyright © 2023 National Sleep Foundation. Published by Elsevier Inc. All rights reserved.)
Databáze: MEDLINE
Externí odkaz: