Improved running gait parameter estimation from single foot-mounted IMU data based on refined event detection

Area of Science:

• Biomechanics
• Wearable Technology
• Sports Science

Background:

• Accurate gait analysis is crucial for portable monitoring using inertial measurement units (IMUs).
• Conventional IMU algorithms struggle with running gait due to speed variations and diverse foot-strike patterns, necessitating adaptive strategies.
• High precision running gait analysis requires improved event detection methods.

Purpose of the Study:

• To introduce MFD-GED (multi-sensor fusion with dynamic gait event detection), a novel method for precise running gait analysis using a single foot-mounted IMU.
• To enhance gait event detection (initial contact, terminal contact, mid-stance) by fusing acceleration and angular velocity data.
• To compute comprehensive running biomechanics parameters and assess the method's validity and performance improvements.

Main Methods:

• The MFD-GED framework fuses acceleration and angular velocity features from a foot-mounted IMU.
• A parametric strategy identifies key gait events: initial contact (IC), terminal contact (TC), and mid-stance (MS).
• The method was validated against a laboratory reference system (optical motion capture, force plates) using correlation coefficients, Bland-Altman analysis, and paired t-tests.

Main Results:

• MFD-GED demonstrated high concurrent validity with the laboratory reference system (Pearson's r = 0.743-0.991, ICC = 0.741-0.990).
• Compared to the angular-velocity-based gait-segmentation (AVGS) method, MFD-GED significantly reduced bias in spatial parameters (e.g., stride velocity, stride length) and temporal parameters (e.g., contact time, flight time).
• Bias in peak vertical ground reaction force (vGRF) also decreased, with reduced error standard deviations across all measured metrics.

Conclusions:

• The MFD-GED framework effectively improves running gait detection and enables high-fidelity parameter estimation using IMUs.
• The method shows significant potential for future gait monitoring applications, offering a reliable tool for professionals.
• While validated in healthy young males, the findings support its utility for advanced running gait analysis.