An Automatic Gait Feature Extraction Method for Identifying Gait Asymmetry Using Wearable Sensors

Area of Science:

• Biomechanics
• Wearable Technology
• Gait Analysis

Background:

• Gait asymmetry is a key indicator of various neurological and musculoskeletal conditions.
• Traditional gait analysis requires specialized laboratory equipment, limiting accessibility.
• Objective, quantitative gait feature extraction is crucial for accurate diagnosis.

Purpose of the Study:

• To assess the efficacy of Inertial Measurement Unit (IMU) sensors for identifying gait asymmetry.
• To develop and validate an Android application for real-time, synchronous IMU data collection from the legs.
• To extract automatic gait features and compare them with a motion capture system.

Main Methods:

• Development of an Android application for collecting synchronized IMU data from both legs.
• Data collection from 20 subjects (10 young, 10 older) performing walking trials.
• Validation of IMU-derived gait features against a Qualisys Motion Capture System.
• Analysis of parameters including distance, time, velocity, stride, step, cadence, and stance/swing phases.

Main Results:

• High accuracy in stride number detection (100% for young, 92.67% for older subjects).
• Accurate estimation of travelled distance (e.g., 97.73% for young, 88.71% for older subjects' right leg).
• Significant differences in average travelled distance and time between young and older subjects, indicating age-related gait changes.

Conclusions:

• Wearable IMU sensors provide a viable, cost-effective alternative to laboratory-based gait analysis.
• The developed system can accurately identify gait asymmetry and diagnose abnormalities.
• This technology facilitates home-based self-assessment of gait, improving accessibility to monitoring and diagnosis.