Physical activity profiling: Activity-specific step counting and energy expenditure models using 3D wrist...

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Human Movement Analysis

Background:

• Accurate physical activity profiling is crucial for health monitoring and research.
• Existing wearable devices often face challenges in precisely measuring energy expenditure and step counts across various activities.
• Developing reliable, wrist-located systems can enhance the accessibility and accuracy of personal health data.

Purpose of the Study:

• To evaluate a novel wrist-located system for physical activity profiling.
• To assess the accuracy of proposed step counting and energy expenditure (EE) estimation methods.
• To validate the system's performance against gold-standard measurements.

Main Methods:

• Utilized a wrist-located sensor for activity profiling.
• Employed an actimetry sensor (waist) and indirect calorimetry for reference data.
• Recruited 13 subjects performing activities like lying, standing, walking, and running at varied intensities.

Main Results:

• The energy expenditure (EE) model achieved a root-mean-square error (RMSE) of 1.22±0.34 kcal/min.
• The step-count model demonstrated an accuracy of 0.71±0.06 steps/10sec during walking and running.
• The system showed high precision in quantifying physical activity parameters.

Conclusions:

• The evaluated wrist-located system provides accurate step counting and energy expenditure estimations.
• This technology offers a reliable solution for unobtrusive physical activity monitoring.
• The findings support the potential of wrist-worn devices for comprehensive health and fitness tracking.