Accelerometer Calibration: The Importance of Considering Functionality | JoVE Visualize

Area of Science:

Biomedical Engineering
Kinesiology
Rehabilitation Science

Background:

Wearable accelerometers are increasingly used to estimate energy expenditure during physical activity.
Accuracy of accelerometer-based energy cost prediction can be influenced by individual factors such as motor performance and underlying health conditions.
Existing prediction models may not adequately account for variations in movement patterns associated with neurological and musculoskeletal diseases.

Purpose of the Study:

To compare the accuracy and precision of hip-worn accelerometers in predicting energy cost (METs) across different motor performance levels and disease conditions.
To evaluate the influence of functional impairment versus specific disease diagnoses on accelerometer-based energy expenditure estimations.
To identify potential improvements for accelerometer calibration to enhance prediction accuracy.

Main Methods:

118 adults, including healthy individuals and those with arthritis, multiple sclerosis, Parkinson's disease, and stroke, were categorized into three functional impairment groups.
Participants performed structured activities while wearing a hip-worn accelerometer and a portable metabolic measurement system.
Accelerometer-predicted energy cost (METs) were compared to measured METs, with statistical analysis using linear mixed effect models and Bayesian information criteria.

Main Results:

Accelerometer counts per minute (CPM) were significantly lower in participants with disease compared to healthy individuals.
Predicted MET bias showed similar patterns across different disease conditions and functional impairment groups.
Model fit was better when considering functional impairment compared to disease condition, suggesting functionality is a key factor.

Conclusions:

Accelerometer calibration using functional impairment levels, rather than disease-specific categories, may reduce prediction variability.
Further research into functionality-based calibration is warranted to improve the accuracy of accelerometer-based physical activity monitoring.
This approach could lead to more precise estimations of energy expenditure in diverse populations.