Self-Supervised Machine Learning to Characterize Step Counts from Wrist-Worn Accelerometers in the UK Biobank.

Journal: Medicine and science in sports and exercise
PMID: 38768076

Abstract

PURPOSE: Step count is an intuitive measure of physical activity frequently quantified in health-related studies; however, accurate step counting is difficult in the free-living environment, with error routinely above 20% in wrist-worn devices against camera-annotated ground truth. This study aimed to describe the development and validation of step count derived from a wrist-worn accelerometer and assess its association with cardiovascular and all-cause mortality in a large prospective cohort.

Authors

  • Scott R Small
  • Shing Chan
  • Rosemary Walmsley
  • Lennart VON Fritsch
    Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Oxford, UNITED KINGDOM.
  • Aidan Acquah
  • Gert Mertes
    Department of Engineering Science, Institute of Biomedical Engineering, University of Oxford, Oxford OX1 2JD, UK.
  • Benjamin G Feakins
  • Andrew Creagh
  • Adam Strange
    SwissRe Institute, UNITED KINGDOM.
  • Charles E Matthews
    Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD.
  • David A Clifton
  • Andrew J Price
    Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Oxford, UNITED KINGDOM.
  • Sara Khalid
    Center for Statistics in Medicine, Botnar Research Center, University of Oxford, Oxford, UK. Electronic address: sara.khalid@ndorms.ox.ac.uk.
  • Derrick Bennett
    Clinical Trial Service Unit and Epidemiological Studies, Nuffield Department of Population Health, University of Oxford, Oxford, UK.
  • Aiden Doherty
    Big Data Institute, Li Ka Shing Centre for Health Information and Discovery, University of Oxford, Oxford, UK. aiden.doherty@bdi.ox.ac.uk.

Keywords

External Resources

Popular Topics
Recent Journals