CoChat
Sign Up
Cited by 41Context-Aware Activity Recognition SystemsStroke Rehabilitation and RecoveryArtificial Intelligence in Healthcare

Measuring activities of daily living in stroke patients with motion machine learning algorithms: A pilot study

Open access
Chen, Pin-Wei, Baune, Nathan A., Zwir, Igor, Wang, Jiayu, Swamidass, Victoria, Wong, Alex W.K.·2021·International Journal of Environmental Research and Public Health
DOI 10.3390/ijerph18041634PMC PMC7915561
Open this paper in CoChat — free accountOpen PDF

Abstract

Measuring activities of daily living (ADLs) using wearable technologies may offer higher precision and granularity than the current clinical assessments for patients after stroke. This study aimed to develop and determine the accuracy of detecting different ADLs using machine-learning (ML) algorithms and wearable sensors. Eleven post-stroke patients participated in this pilot study at an ADL Simulation Lab across two study visits. We collected blocks of repeated activity ("atomic" activity) performance data to train our ML algorithms during one visit. We evaluated our ML algorithms using independent semi-naturalistic activity data collected at a separate session. We tested Decision Tree, Random Forest, Support Vector Machine (SVM), and eXtreme Gradient Boosting (XGBoost) for model development. XGBoost was the best classification model. We achieved 82% accuracy based on ten ADL tasks. With a model including seven tasks, accuracy improved to 90%. ADL tasks included chopping food, vacuuming, sweeping, spreading jam or butter, folding laundry, eating, brushing teeth, taking off/putting on a shirt, wiping a cupboard, and buttoning a shirt. Results provide preliminary evidence that ADL functioning can be predicted with adequate accuracy using wearable sensors and ML. The use of external validation (independent training and testing data sets) and semi-naturalistic testing data is a major strength of the study and a step closer to the long-term goal of ADL monitoring in real-world settings. Further investigation is needed to improve the ADL prediction accuracy, increase the number of tasks monitored, and test the model outside of a laboratory setting.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.3390/ijerph18041634

Cites

40 of 40 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • In study [49], the best performance among the evaluated models was 82% accuracy.

    “82%”

    From Wearable sensors and machine learning in post-stroke rehabilitation assessment: A systematic review (2022) · cites this paper as [49]

    The Results (§3.2) report 82% accuracy on the independent test set for the ten-task model, but 90% for the seven-task model. Thus 82% was a reported result, not the best accuracy reported in the paper. (LLM, conf=0.91)

    Checked by a reasoning model — may be wrong.

    Experimental — accuracy not yet measured

  • Data bias causing assessment algorithms not to generalise to new patients with different degrees of severity was very common among the reviewed papers, with researchers complaining about poor generalisation.

    “In the same context, data bias which is another issue can cause the assessment algorithm to evolve in an unpredictable manner and not generalise to new patients that have different degrees of severity.This was very common among the reviewed papers, where researchers complained about their algorithms not generalising well.[41,49,52,58,60].”

    From Wearable sensors and machine learning in post-stroke rehabilitation assessment: A systematic review (2022) · cites this paper as [49]

    The Discussion says models trained on mild stroke patients may not generalise to patients with moderate or severe stroke, and calls generalisability a major concern. It does not establish that poor generalisation was very common among reviewed papers, that researchers reported their algorithms failing to generalise, or that data bias made algorithms evolve unpredictably. (LLM, conf=0.86)

    Experimental — accuracy not yet measured

  • Designing efficient ML-based assessment systems that generalise to new users requires a large number of participants with different and variant degrees of severity.

    “In addition, in order to design efficient ML-based assessment systems that generalises to new users a large number of participants should take part with different and variant degrees of severity [43,48,49,69,70], which is not always available and taken into account.”

    From Wearable sensors and machine learning in post-stroke rehabilitation assessment: A systematic review (2022) · cites this paper as [49]

    The Discussion identifies the small sample as a limitation and calls for larger-scale data collection. It also warns that models trained on mild stroke patients may not generalize to moderate or severe cases. However, it does not establish that a large, severity-diverse participant group is required for a model that generalizes to new users; it leaves open whether one model or separate models would work across severity groups. (LLM, conf=0.82)

    Experimental — accuracy not yet measured

  • In activity recognition, researchers also covered kitchen tasks and routine activities such as making a drink, chopping food, making the bed, reading, lacing shoes, folding, sweeping, and brushing teeth.

    “performing kitchen tasks like making a drink, chopping food [42] and other routine activities like making the bed, reading and lacing shoes [48], folding, sweeping and brushing teeth [46,48,49].”

    From Wearable sensors and machine learning in post-stroke rehabilitation assessment: A systematic review (2022) · cites this paper as [49]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Chen, P.-W., Baune, N. A., Zwir, I., Wang, J., Swamidass, V., & Wong, A. W. K.
Affiliations
  • Washington University in St. Louis (US)
  • Universidad de Granada (ES)
  • PlatformSTL, St. Louis, MO 63110, USA
  • Shirley Ryan AbilityLab (US)
Corresponding author
Wong, Alex W.K.
ORCID
Venue
International Journal of Environmental Research and Public Health
Publisher
MDPI AG
Published
2021-02-09
Volume
18
Issue
4
Pages
1634
ISSN
1660-4601 (ISSN-L)
DOI
10.3390/ijerph18041634
PMC
PMC7915561
OpenAlex
W3126701990
Cited by
41
References
40
Citations by year
2026: 6 · 2025: 3 · 2024: 8 · 2023: 12 · 2022: 8 · 2021: 4
FWCI
2.88
Type
journal article
Language
en
Open access
Licence
Locations
  • International Journal of Environmental Research and Public Health (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Digital Commons@Becker (Washington University School of Medicine) (submittedVersion)
  • Institutional Repository of the University of Granada (University of Granada) (submittedVersion, cc-by-nc-nd)
  • RePEc: Research Papers in Economics (submittedVersion)
  • Institutional Repository of the University of Granada (University of Granada) (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Funders
  • National Institute on Disability, Independent Living, and Rehabilitation Research (90BISA0015)
  • National Center for Medical Rehabilitation Research (K01HD095388)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
  • Health Sciences › Medicine › Rehabilitation › Stroke Rehabilitation and Recovery
  • Health Sciences › Health Professions › Health Information Management › Artificial Intelligence in Healthcare
Keywords
activities of daily living, stroke, wearable sensors, motion sensing, machine learning classification, XGBoost, stroke rehabilitation, activity recognition, external validation
MeSH terms
Machine Learning, Activities of Daily Living, Algorithms, Humans, Pilot Projects, Stroke

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review