Measuring Walking Speed Using Actibelt
Author Information
Author(s): Schimpl Michaela, Lederer Christian, Daumer Martin, Villoslada Pablo
Primary Institution: Trium Analysis Online GmbH, Munich, Germany
Hypothesis
Can walking speed be accurately measured in free-living environments using mobile accelerometry?
Conclusion
The study found that a novel algorithm using support vector regression can accurately measure walking speed in free-living environments.
Supporting Evidence
- A novel algorithm employing support vector regression was found to perform best with a concordance correlation coefficient of 0.93.
- The study included 17 healthy subjects to validate the walking speed measurement method.
- Coverage probability for a deviation of 0.1 m/s was found to be 0.46.
Takeaway
This study shows a new way to measure how fast people walk in their everyday lives using a special device that fits around their waist.
Methodology
The study developed algorithms for walking speed prediction based on 3D accelerometry data from the Actibelt® and validated these against a mobile gold standard.
Potential Biases
Potential bias due to the calibration of algorithms based on specific walking tests.
Limitations
The mobile gold standard may constrain natural arm swing and boundary effects can occur during starting or stopping.
Participant Demographics
17 healthy subjects (41% male, aged 32 ± 15 years).
Statistical Information
P-Value
0.93
Confidence Interval
95%CI 0.92–0.94
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website