A novel algorithm to detect non-wear time from raw accelerometer data using deep convolutional neural networks

Abstract To date, non-wear detection algorithms commonly employ a 30, 60, or even 90 mins interval or window in which acceleration values need to be below a threshold value. A major drawback of such intervals is that they need to be long enough to prevent false positives (type I errors), while short...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Shaheen Syed, Bente Morseth, Laila A. Hopstock, Alexander Horsch
Formato: article
Lenguaje:EN
Publicado: Nature Portfolio 2021
Materias:
R
Q
Acceso en línea:https://doaj.org/article/23d4d98bfba6484ebb75713885ef5f1e
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:23d4d98bfba6484ebb75713885ef5f1e
record_format dspace
spelling oai:doaj.org-article:23d4d98bfba6484ebb75713885ef5f1e2021-12-02T16:45:15ZA novel algorithm to detect non-wear time from raw accelerometer data using deep convolutional neural networks10.1038/s41598-021-87757-z2045-2322https://doaj.org/article/23d4d98bfba6484ebb75713885ef5f1e2021-04-01T00:00:00Zhttps://doi.org/10.1038/s41598-021-87757-zhttps://doaj.org/toc/2045-2322Abstract To date, non-wear detection algorithms commonly employ a 30, 60, or even 90 mins interval or window in which acceleration values need to be below a threshold value. A major drawback of such intervals is that they need to be long enough to prevent false positives (type I errors), while short enough to prevent false negatives (type II errors), which limits detecting both short and longer episodes of non-wear time. In this paper, we propose a novel non-wear detection algorithm that eliminates the need for an interval. Rather than inspecting acceleration within intervals, we explore acceleration right before and right after an episode of non-wear time. We trained a deep convolutional neural network that was able to infer non-wear time by detecting when the accelerometer was removed and when it was placed back on again. We evaluate our algorithm against several baseline and existing non-wear algorithms, and our algorithm achieves a perfect precision, a recall of 0.9962, and an F1 score of 0.9981, outperforming all evaluated algorithms. Although our algorithm was developed using patterns learned from a hip-worn accelerometer, we propose algorithmic steps that can easily be applied to a wrist-worn accelerometer and a retrained classification model.Shaheen SyedBente MorsethLaila A. HopstockAlexander HorschNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-12 (2021)
institution DOAJ
collection DOAJ
language EN
topic Medicine
R
Science
Q
spellingShingle Medicine
R
Science
Q
Shaheen Syed
Bente Morseth
Laila A. Hopstock
Alexander Horsch
A novel algorithm to detect non-wear time from raw accelerometer data using deep convolutional neural networks
description Abstract To date, non-wear detection algorithms commonly employ a 30, 60, or even 90 mins interval or window in which acceleration values need to be below a threshold value. A major drawback of such intervals is that they need to be long enough to prevent false positives (type I errors), while short enough to prevent false negatives (type II errors), which limits detecting both short and longer episodes of non-wear time. In this paper, we propose a novel non-wear detection algorithm that eliminates the need for an interval. Rather than inspecting acceleration within intervals, we explore acceleration right before and right after an episode of non-wear time. We trained a deep convolutional neural network that was able to infer non-wear time by detecting when the accelerometer was removed and when it was placed back on again. We evaluate our algorithm against several baseline and existing non-wear algorithms, and our algorithm achieves a perfect precision, a recall of 0.9962, and an F1 score of 0.9981, outperforming all evaluated algorithms. Although our algorithm was developed using patterns learned from a hip-worn accelerometer, we propose algorithmic steps that can easily be applied to a wrist-worn accelerometer and a retrained classification model.
format article
author Shaheen Syed
Bente Morseth
Laila A. Hopstock
Alexander Horsch
author_facet Shaheen Syed
Bente Morseth
Laila A. Hopstock
Alexander Horsch
author_sort Shaheen Syed
title A novel algorithm to detect non-wear time from raw accelerometer data using deep convolutional neural networks
title_short A novel algorithm to detect non-wear time from raw accelerometer data using deep convolutional neural networks
title_full A novel algorithm to detect non-wear time from raw accelerometer data using deep convolutional neural networks
title_fullStr A novel algorithm to detect non-wear time from raw accelerometer data using deep convolutional neural networks
title_full_unstemmed A novel algorithm to detect non-wear time from raw accelerometer data using deep convolutional neural networks
title_sort novel algorithm to detect non-wear time from raw accelerometer data using deep convolutional neural networks
publisher Nature Portfolio
publishDate 2021
url https://doaj.org/article/23d4d98bfba6484ebb75713885ef5f1e
work_keys_str_mv AT shaheensyed anovelalgorithmtodetectnonweartimefromrawaccelerometerdatausingdeepconvolutionalneuralnetworks
AT bentemorseth anovelalgorithmtodetectnonweartimefromrawaccelerometerdatausingdeepconvolutionalneuralnetworks
AT lailaahopstock anovelalgorithmtodetectnonweartimefromrawaccelerometerdatausingdeepconvolutionalneuralnetworks
AT alexanderhorsch anovelalgorithmtodetectnonweartimefromrawaccelerometerdatausingdeepconvolutionalneuralnetworks
AT shaheensyed novelalgorithmtodetectnonweartimefromrawaccelerometerdatausingdeepconvolutionalneuralnetworks
AT bentemorseth novelalgorithmtodetectnonweartimefromrawaccelerometerdatausingdeepconvolutionalneuralnetworks
AT lailaahopstock novelalgorithmtodetectnonweartimefromrawaccelerometerdatausingdeepconvolutionalneuralnetworks
AT alexanderhorsch novelalgorithmtodetectnonweartimefromrawaccelerometerdatausingdeepconvolutionalneuralnetworks
_version_ 1718383443412254720