Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Detection of unilateral arm paresis after stroke by wearable accelerometers and machine learning

Wasselius, Johan LU ; Finn, Eric Lyckegård ; Persson, Emma ; Ericson, Petter ; Brogårdh, Christina LU ; Lindgren, Arne G. LU ; Ullberg, Teresa LU and Åström, Kalle LU orcid (2021) In Sensors 21(23).
Abstract

Recent advances in stroke treatment have provided effective tools to successfully treat ischemic stroke, but still a majority of patients are not treated due to late arrival to hospital. With modern stroke treatment, earlier arrival would greatly improve the overall treatment results. This prospective study was performed to asses the capability of bilateral accelerometers worn in bracelets 24/7 to detect unilateral arm paralysis, a hallmark symptom of stroke, early enough to receive treatment. Classical machine learning algorithms as well as state-of-the-art deep neural networks were evaluated on detection times between 15 min and 120 min. Motion data were collected using triaxial accelerometer bracelets worn on both arms for 24 h.... (More)

Recent advances in stroke treatment have provided effective tools to successfully treat ischemic stroke, but still a majority of patients are not treated due to late arrival to hospital. With modern stroke treatment, earlier arrival would greatly improve the overall treatment results. This prospective study was performed to asses the capability of bilateral accelerometers worn in bracelets 24/7 to detect unilateral arm paralysis, a hallmark symptom of stroke, early enough to receive treatment. Classical machine learning algorithms as well as state-of-the-art deep neural networks were evaluated on detection times between 15 min and 120 min. Motion data were collected using triaxial accelerometer bracelets worn on both arms for 24 h. Eighty-four stroke patients with unilateral arm motor impairment and 101 healthy subjects participated in the study. Accelerometer data were divided into data windows of different lengths and analyzed using multiple machine learning algorithms. The results show that all algorithms performed well in separating the two groups early enough to be clinically relevant, based on wrist-worn accelerometers. The two evaluated deep learning models, fully convolutional network and InceptionTime, performed better than the classical machine learning models with an AUC score between 0.947–0.957 on 15 min data windows and up to 0.993–0.994 on 120 min data windows. Window lengths longer than 90 min only marginally improved performance. The difference in performance between the deep learning models and the classical models was statistically significant according to a non-parametric Friedman test followed by a post-hoc Nemenyi test. Introduction of wearable stroke detection devices may dramatically increase the portion of stroke patients eligible for revascularization and shorten the time to treatment. Since the treatment effect is highly time-dependent, early stroke detection may dramatically improve stroke outcomes.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Accelerometer, Artificial intelligence, Hemiparesis, Machine learning, Monitoring, Motor deficit, Sensor, Stroke, Wearable
in
Sensors
volume
21
issue
23
article number
7784
publisher
MDPI AG
external identifiers
  • scopus:85119589014
  • pmid:34883800
ISSN
1424-8220
DOI
10.3390/s21237784
language
English
LU publication?
yes
additional info
Funding Information: Funding: This research was funded by THE CRAFOORD FOUNDATION #20180610, THE SWEDISH ALF #I-ALF 47447 and #YF-ALF 43435, SKANE UNIVERSITY HOSPITAL RESEARCH FUNDS #96437 and #96438 to J.W., the Swedish Research Council (2019-01757), CaNVAS project funded by NIH (US), the Swedish Government (under the Avtal om Läkarutbildning och Medicinsk Forskning, ALF), the Swedish Heart and Lung Foundation, Region Skåne, Lund University, Skåne University Hospital, Sparbanksstiftelsen Färs och Frosta, Fremasons Lodge of Instruction Eos in Lund to A.G.L., the strategic research projects ELLIIT and eSSENCE, the Swedish Foundation for Strategic Research project and the Wallenberg Artificial Intelligence, Autonomous Systems, and Software Program (WASP) to K.Å. Publisher Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland.
id
e927a43f-0dca-4476-879d-a4f812c00c1f
date added to LUP
2021-11-29 08:04:35
date last changed
2024-06-15 21:42:58
@article{e927a43f-0dca-4476-879d-a4f812c00c1f,
  abstract     = {{<p>Recent advances in stroke treatment have provided effective tools to successfully treat ischemic stroke, but still a majority of patients are not treated due to late arrival to hospital. With modern stroke treatment, earlier arrival would greatly improve the overall treatment results. This prospective study was performed to asses the capability of bilateral accelerometers worn in bracelets 24/7 to detect unilateral arm paralysis, a hallmark symptom of stroke, early enough to receive treatment. Classical machine learning algorithms as well as state-of-the-art deep neural networks were evaluated on detection times between 15 min and 120 min. Motion data were collected using triaxial accelerometer bracelets worn on both arms for 24 h. Eighty-four stroke patients with unilateral arm motor impairment and 101 healthy subjects participated in the study. Accelerometer data were divided into data windows of different lengths and analyzed using multiple machine learning algorithms. The results show that all algorithms performed well in separating the two groups early enough to be clinically relevant, based on wrist-worn accelerometers. The two evaluated deep learning models, fully convolutional network and InceptionTime, performed better than the classical machine learning models with an AUC score between 0.947–0.957 on 15 min data windows and up to 0.993–0.994 on 120 min data windows. Window lengths longer than 90 min only marginally improved performance. The difference in performance between the deep learning models and the classical models was statistically significant according to a non-parametric Friedman test followed by a post-hoc Nemenyi test. Introduction of wearable stroke detection devices may dramatically increase the portion of stroke patients eligible for revascularization and shorten the time to treatment. Since the treatment effect is highly time-dependent, early stroke detection may dramatically improve stroke outcomes.</p>}},
  author       = {{Wasselius, Johan and Finn, Eric Lyckegård and Persson, Emma and Ericson, Petter and Brogårdh, Christina and Lindgren, Arne G. and Ullberg, Teresa and Åström, Kalle}},
  issn         = {{1424-8220}},
  keywords     = {{Accelerometer; Artificial intelligence; Hemiparesis; Machine learning; Monitoring; Motor deficit; Sensor; Stroke; Wearable}},
  language     = {{eng}},
  number       = {{23}},
  publisher    = {{MDPI AG}},
  series       = {{Sensors}},
  title        = {{Detection of unilateral arm paresis after stroke by wearable accelerometers and machine learning}},
  url          = {{http://dx.doi.org/10.3390/s21237784}},
  doi          = {{10.3390/s21237784}},
  volume       = {{21}},
  year         = {{2021}},
}