Real-time gait analysis with accelerometer-based smart shoes.

Area of Science:

• Biomechanics
• Wearable Technology
• Sensor Systems

Background:

• Gait analysis is crucial for understanding human locomotion and detecting abnormalities.
• Existing gait analysis systems often require specialized equipment or are limited in real-time application.
• Smart footwear offers a promising avenue for unobtrusive and continuous gait monitoring.

Purpose of the Study:

• To evaluate a novel smart shoe system for real-time gait analysis.
• To assess the system's ability to estimate key gait parameters using only accelerometers.
• To validate the accuracy and precision of the smart shoe's measurements against a gold standard.

Main Methods:

• Development of a smart shoe system utilizing solely accelerometers for reduced power consumption.
• Real-time estimation of gait parameters including activity class, step cadence, ground contact time, foot impact characteristics, distance, and speed.
• Validation using a custom database of 26 subjects on a treadmill, with manual video analysis as a reference.

Main Results:

• High accuracy in retrieving running cadence (max 2% error) and ground contact time (avg 3.25% error).
• Precise classification of foot impact zones, achieving 72% to 91% precision depending on running style.
• Successful real-time estimation of various gait parameters with minimal sensor components.

Conclusions:

• The developed smart shoe system provides an effective and low-power solution for real-time gait analysis.
• Accelerometer-based gait analysis in footwear is a viable technology for diverse applications.
• The algorithm's successful licensing indicates its commercial potential in the wearable technology market.