OpenMetabolics: Estimating energy expenditure using a smartphone worn in a pocket

Area of Science:

• Biomedical Engineering
• Health Informatics
• Wearable Technology

Background:

• Physical inactivity is a major global mortality cause, necessitating objective physical activity measurement tools.
• Current activity monitors have limitations in accuracy and suitability for large-scale studies.
• Understanding activity-health outcome relationships requires improved monitoring methods.

Purpose of the Study:

• To develop and validate an accurate, accessible physical activity monitor using smartphone technology.
• To improve the estimation of energy expenditure during daily activities.
• To facilitate large-scale research on physical activity and health outcomes.

Main Methods:

• Developed OpenMetabolics, a biomechanically-informed system using a smartphone in a pocket to measure leg motion.
• Employed a data-driven machine learning model to correlate leg muscle activity with energy expenditure.
• Implemented a pocket motion artifact correction model for accurate monitoring across different clothing types.

Main Results:

• OpenMetabolics achieved 18% cumulative error in energy expenditure estimation across real-world activities, outperforming existing tools by approximately two times.
• The pocket motion artifact correction model ensured accurate monitoring regardless of clothing.
• A week-long study revealed individual and population-level activity patterns.

Conclusions:

• OpenMetabolics offers a significant advancement in accurate and accessible physical activity monitoring.
• The open-source nature of the system promotes widespread adoption for research and public health initiatives.
• This tool can support large-scale studies to link physical activity with health outcomes, inform policy, and guide interventions.