Walking and running cadence estimation using a single trunk-fixed accelerometer for daily physical activities...

Area of Science:

• Biomedical Engineering
• Sports Science
• Wearable Technology

Background:

• Accurate physical activity (PA) assessment is crucial for understanding health and well-being.
• Lightweight wearable sensors are increasingly used for PA monitoring.
• Activity monitors can evaluate the health impacts of PA interventions.

Purpose of the Study:

• To evaluate the acute and long-term effects of an eight-week running intervention on PA behavior and vitality.
• To develop and validate algorithms for detecting running and estimating cadence using a single trunk-fixed accelerometer.

Main Methods:

• Developed algorithms for running detection and cadence estimation (time and frequency domains) using trunk-fixed accelerometer data.
• Validated algorithms against an open-source gait database across various locomotion speeds.
• The Run4Vit study utilized these algorithms to assess PA changes.

Main Results:

• Cadence estimation algorithms showed low bias and high precision (-0.01 ± 0.69 steps/min for temporal, 0.02 ± 1.33 steps/min for frequency).
• The running detection algorithm achieved 98% accuracy and >99% precision.
• Validated algorithms provide reliable metrics for PA dimensions.

Conclusions:

• The developed algorithms accurately assess running and cadence, crucial for PA monitoring.
• These validated tools can provide reliable outcome measures for longitudinal PA interventions like Run4Vit.
• This technology supports multi-dimensional PA classification for health and vitality research.