Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Development of novel techniques to classify physical activity mode using accelerometers.

David M Pober1, John Staudenmayer, Christopher Raphael

  • 1Department of Exercise Science, Exercise Physiology Laboratory, University of Massachusetts, Amherst, MA 01003, USA.

Medicine and Science in Sports and Exercise
|September 9, 2006
PubMed
Summary

Sophisticated data processing for accelerometers can accurately distinguish between different physical activities (PA). This improved method provides a more precise assessment of PA compared to traditional techniques.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

An Examination of Efficiency during Walking in Children and Adults.

Pediatric exercise science·2025
Same author

Feasibility and Costs of Monitoring Physical Activity in Young Children Using the Caltrac Accelerometer.

Pediatric exercise science·2025
Same author

Validity of the Caltrac Accelerometer in Estimating Energy Expenditure and Activity in Children and Adults.

Pediatric exercise science·2024
Same author

Validity of the Pregnancy Physical Activity Questionnaire Short Form (PPAQ-SF).

American journal of epidemiology·2024
Same author

Cadence (steps/min) and relative intensity in 61 to 85-year-olds: the CADENCE-Adults study.

The international journal of behavioral nutrition and physical activity·2023
Same author

Cadence (steps/min) as an indicator of the walk-to-run transition.

Human movement science·2023

Area of Science:

  • Biomedical Engineering
  • Public Health
  • Wearable Technology

Background:

  • Accelerometers are widely used in public health research to measure physical activity (PA).
  • Current data processing techniques often limit the accuracy of accelerometer-based PA assessments.
  • Developing more sophisticated methods is crucial for reliable PA monitoring.

Purpose of the Study:

  • To investigate if advanced data processing can differentiate between various activity types using accelerometer data.
  • To enhance the accuracy of physical activity assessment in public health research.

Main Methods:

  • Utilized MTI Actigraphs worn by six subjects during four distinct activities: walking, walking uphill, vacuuming, and computer work.
  • Applied quadratic discriminant analysis (QDA) and trained a hidden Markov model (HMM) to recognize these activities.

Related Experiment Videos

  • Assessed the classification accuracy of these novel analytical techniques.

    • Main Results:

    • The hidden Markov model (HMM) achieved 80.8% accuracy in identifying activity modes, outperforming quadratic discriminant analysis (QDA) at 70.9%.
    • HMM demonstrated high accuracy in recognizing vacuuming (98.8%) and computer work (97.3%).
    • QDA and HMM correctly estimated activity intensity 99% of the time, contrasting with traditional methods that misidentified 100% of vacuuming and uphill walking time.

    Conclusions:

    • Estimating activity mode, rather than just activity level, using accelerometer data offers a more accurate method for field-based PA assessment.
    • This novel approach shows promise for improving PA measurement accuracy.
    • Further research with larger, diverse populations and activities is recommended to validate these findings.