Comparing adaptive algorithms to measure temporal gait parameters using lower body mounted inertial sensors

Area of Science:

• Biomechanics
• Wearable Technology
• Gait Analysis

Background:

• Accurate temporal gait parameters are crucial for assessing mobility and diagnosing conditions.
• Inertial Measurement Units (IMUs) offer a portable solution for gait data acquisition.
• Adaptive algorithms are needed to process IMU data effectively for gait event detection.

Purpose of the Study:

• To compare the accuracy of three adaptive algorithms (Greene, Selles, Sabatini) in calculating temporal gait parameters.
• To evaluate algorithm performance using data from IMUs worn on the feet and shanks.
• To validate algorithm-derived parameters against a gold standard force plate system.

Main Methods:

• Eight subjects completed walking trials under five conditions (normal, fast, slow, stiff ankle, stiff knee).
• IMU data from shank and foot sensors were processed using Greene, Selles, and Sabatini adaptive algorithms.
• Temporal gait parameters (stance time, double support time) were calculated and compared to force plate measurements.

Main Results:

• The Greene method demonstrated high accuracy for stance time (r=0.990) and double support time (r=0.837).
• Sabatini algorithm showed good accuracy for stance time (r=0.980) but moderate for double support (r=0.745).
• Selles algorithm exhibited lower accuracy for both stance time (r=0.865) and double support time (r=0.583).

Conclusions:

• The Greene adaptive algorithm is the most accurate method for estimating temporal gait events from IMU data.
• This finding supports the use of the Greene algorithm in wearable sensor-based gait analysis.
• Accurate gait parameter estimation using IMUs has implications for clinical and research applications.