Personalized Sleep Parameters Estimation from Actigraphy: A Machine Learning Approach

Aria Khademi,1–3 Yasser EL-Manzalawy,1,4 Lindsay Master,5 Orfeu M Buxton,5–9 Vasant G Honavar1–3,6,10,11 1College of Information Sciences and Technology, The Pennsylvania State University, University Park, PA, USA; 2Artificial Intelligence Research Laboratory, The Penns...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Khademi A, EL-Manzalawy Y, Master L, Buxton OM, Honavar VG
Formato: article
Lenguaje:EN
Publicado: Dove Medical Press 2019
Materias:
Acceso en línea:https://doaj.org/article/0a9fdc6c6b5e473db580ad3dc6c74aaf
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:0a9fdc6c6b5e473db580ad3dc6c74aaf
record_format dspace
institution DOAJ
collection DOAJ
language EN
topic actigraphy
polysomnography
personalized
machine learning
sleep parameters
Psychiatry
RC435-571
Neurophysiology and neuropsychology
QP351-495
spellingShingle actigraphy
polysomnography
personalized
machine learning
sleep parameters
Psychiatry
RC435-571
Neurophysiology and neuropsychology
QP351-495
Khademi A
EL-Manzalawy Y
Master L
Buxton OM
Honavar VG
Personalized Sleep Parameters Estimation from Actigraphy: A Machine Learning Approach
description Aria Khademi,1–3 Yasser EL-Manzalawy,1,4 Lindsay Master,5 Orfeu M Buxton,5–9 Vasant G Honavar1–3,6,10,11 1College of Information Sciences and Technology, The Pennsylvania State University, University Park, PA, USA; 2Artificial Intelligence Research Laboratory, The Pennsylvania State University, University Park, PA, USA; 3Center for Big Data Analytics and Discovery Informatics, The Pennsylvania State University, University Park, PA, USA; 4Department of Imaging Science and Innovation, Geisinger Health System, Danville, PA, 17822, USA; 5Department of Biobehavioral Health, The Pennsylvania State University, University Park, PA, USA; 6Clinical and Translational Sciences Institute, The Pennsylvania State University, University Park, PA, USA; 7Division of Sleep Medicine, Harvard University, Boston, MA, USA; 8Department of Social and Behavioral Sciences, Harvard Chan School of Public Health, Boston, MA, USA; 9Division of Sleep and Circadian Disorders, Departments of Medicine and Neurology, Brigham and Women’s Hospital, Boston, MA, USA; 10Institute for Computational and Data Sciences, The Pennsylvania State University, University Park, PA, USA; 11Huck Institutes of the Life Sciences, The Pennsylvania State University, University Park, PA, USACorrespondence: Orfeu M BuxtonThe Pennsylvania State University, University Park, PA 16802, USATel +1 814 865 3141Email orfeu@psu.eduBackground: The current gold standard for measuring sleep is polysomnography (PSG), but it can be obtrusive and costly. Actigraphy is a relatively low-cost and unobtrusive alternative to PSG. Of particular interest in measuring sleep from actigraphy is prediction of sleep-wake states. Current literature on prediction of sleep-wake states from actigraphy consists of methods that use population data, which we call generalized models. However, accounting for variability of sleep patterns across individuals calls for personalized models of sleep-wake states prediction that could be potentially better suited to individual-level data and yield more accurate estimation of sleep.Purpose: To investigate the validity of developing personalized machine learning models, trained and tested on individual-level actigraphy data, for improved prediction of sleep-wake states and reliable estimation of nightly sleep parameters.Participants and methods: We used a dataset including 54 participants and systematically trained and tested 5 different personalized machine learning models as well as their generalized counterparts. We evaluated model performance compared to concurrent PSG through extensive machine learning experiments and statistical analyses.Results: Our experiments show the superiority of personalized models over their generalized counterparts in estimating PSG-derived sleep parameters. Personalized models of regularized logistic regression, random forest, adaptive boosting, and extreme gradient boosting achieve estimates of total sleep time, wake after sleep onset, sleep efficiency, and number of awakenings that are closer to those obtained by PSG, in absolute difference, than the same estimates from their generalized counterparts. We further show that the difference between estimates of sleep parameters obtained by personalized models and those of PSG is statistically non-significant.Conclusion: Personalized machine learning models of sleep-wake states outperform their generalized counterparts in terms of estimating sleep parameters and are indistinguishable from PSG labeled sleep-wake states. Personalized machine learning models can be used in actigraphy studies of sleep health and potentially screening for some sleep disorders.Keywords: actigraphy, polysomnography, personalized, machine learning, sleep parameters
format article
author Khademi A
EL-Manzalawy Y
Master L
Buxton OM
Honavar VG
author_facet Khademi A
EL-Manzalawy Y
Master L
Buxton OM
Honavar VG
author_sort Khademi A
title Personalized Sleep Parameters Estimation from Actigraphy: A Machine Learning Approach
title_short Personalized Sleep Parameters Estimation from Actigraphy: A Machine Learning Approach
title_full Personalized Sleep Parameters Estimation from Actigraphy: A Machine Learning Approach
title_fullStr Personalized Sleep Parameters Estimation from Actigraphy: A Machine Learning Approach
title_full_unstemmed Personalized Sleep Parameters Estimation from Actigraphy: A Machine Learning Approach
title_sort personalized sleep parameters estimation from actigraphy: a machine learning approach
publisher Dove Medical Press
publishDate 2019
url https://doaj.org/article/0a9fdc6c6b5e473db580ad3dc6c74aaf
work_keys_str_mv AT khademia personalizedsleepparametersestimationfromactigraphyamachinelearningapproach
AT elmanzalawyy personalizedsleepparametersestimationfromactigraphyamachinelearningapproach
AT masterl personalizedsleepparametersestimationfromactigraphyamachinelearningapproach
AT buxtonom personalizedsleepparametersestimationfromactigraphyamachinelearningapproach
AT honavarvg personalizedsleepparametersestimationfromactigraphyamachinelearningapproach
_version_ 1718401712071376896
spelling oai:doaj.org-article:0a9fdc6c6b5e473db580ad3dc6c74aaf2021-12-02T03:25:57ZPersonalized Sleep Parameters Estimation from Actigraphy: A Machine Learning Approach1179-1608https://doaj.org/article/0a9fdc6c6b5e473db580ad3dc6c74aaf2019-12-01T00:00:00Zhttps://www.dovepress.com/personalized-sleep-parameters-estimation-from-actigraphy-a-machine-lea-peer-reviewed-article-NSShttps://doaj.org/toc/1179-1608Aria Khademi,1–3 Yasser EL-Manzalawy,1,4 Lindsay Master,5 Orfeu M Buxton,5–9 Vasant G Honavar1–3,6,10,11 1College of Information Sciences and Technology, The Pennsylvania State University, University Park, PA, USA; 2Artificial Intelligence Research Laboratory, The Pennsylvania State University, University Park, PA, USA; 3Center for Big Data Analytics and Discovery Informatics, The Pennsylvania State University, University Park, PA, USA; 4Department of Imaging Science and Innovation, Geisinger Health System, Danville, PA, 17822, USA; 5Department of Biobehavioral Health, The Pennsylvania State University, University Park, PA, USA; 6Clinical and Translational Sciences Institute, The Pennsylvania State University, University Park, PA, USA; 7Division of Sleep Medicine, Harvard University, Boston, MA, USA; 8Department of Social and Behavioral Sciences, Harvard Chan School of Public Health, Boston, MA, USA; 9Division of Sleep and Circadian Disorders, Departments of Medicine and Neurology, Brigham and Women’s Hospital, Boston, MA, USA; 10Institute for Computational and Data Sciences, The Pennsylvania State University, University Park, PA, USA; 11Huck Institutes of the Life Sciences, The Pennsylvania State University, University Park, PA, USACorrespondence: Orfeu M BuxtonThe Pennsylvania State University, University Park, PA 16802, USATel +1 814 865 3141Email orfeu@psu.eduBackground: The current gold standard for measuring sleep is polysomnography (PSG), but it can be obtrusive and costly. Actigraphy is a relatively low-cost and unobtrusive alternative to PSG. Of particular interest in measuring sleep from actigraphy is prediction of sleep-wake states. Current literature on prediction of sleep-wake states from actigraphy consists of methods that use population data, which we call generalized models. However, accounting for variability of sleep patterns across individuals calls for personalized models of sleep-wake states prediction that could be potentially better suited to individual-level data and yield more accurate estimation of sleep.Purpose: To investigate the validity of developing personalized machine learning models, trained and tested on individual-level actigraphy data, for improved prediction of sleep-wake states and reliable estimation of nightly sleep parameters.Participants and methods: We used a dataset including 54 participants and systematically trained and tested 5 different personalized machine learning models as well as their generalized counterparts. We evaluated model performance compared to concurrent PSG through extensive machine learning experiments and statistical analyses.Results: Our experiments show the superiority of personalized models over their generalized counterparts in estimating PSG-derived sleep parameters. Personalized models of regularized logistic regression, random forest, adaptive boosting, and extreme gradient boosting achieve estimates of total sleep time, wake after sleep onset, sleep efficiency, and number of awakenings that are closer to those obtained by PSG, in absolute difference, than the same estimates from their generalized counterparts. We further show that the difference between estimates of sleep parameters obtained by personalized models and those of PSG is statistically non-significant.Conclusion: Personalized machine learning models of sleep-wake states outperform their generalized counterparts in terms of estimating sleep parameters and are indistinguishable from PSG labeled sleep-wake states. Personalized machine learning models can be used in actigraphy studies of sleep health and potentially screening for some sleep disorders.Keywords: actigraphy, polysomnography, personalized, machine learning, sleep parametersKhademi AEL-Manzalawy YMaster LBuxton OMHonavar VGDove Medical Pressarticleactigraphypolysomnographypersonalizedmachine learningsleep parametersPsychiatryRC435-571Neurophysiology and neuropsychologyQP351-495ENNature and Science of Sleep, Vol Volume 11, Pp 387-399 (2019)