The implementation of inertial sensors for the assessment of temporal parameters of gait in the knee arthroplasty population

Background: The use of inertial measurement units for the evaluation of temporal parameters of gait has been studied in many populations. However, currently no studies support the use of inertial measurement units for this purpose in the knee arthroplasty population. The objective of the present study was to investigate the agreement between an inertial measurement and camera based system for the assessment of temporal gait parameters in a knee arthroplasty population.

Methods: Sixteen knee arthroplasty patients performed 3 gait trials at a self-selected speed along a 6 m walk-way. During the gait trials, gyroscope data from shank-worn inertial measurement units and motion data from optoelectronic cameras were collected simultaneously. A custom-made peak detection algorithm was used to identify gait events from gyroscope data, in order to compute cycle time, stance time and swing time. A marker and coordinate based algorithm was used to calculate temporal gait parameters from kinematical data derived from the camera system. Temporal variables were compared between both methods by calculating intra-class correlation coefficients, mean errors and root mean squared errors. Furthermore, Bland-Altman plots were constructed to assess the agreement between both methods.

Findings: Overall good to excellent intra-class correlation values (0.826-0.972) were found. Root mean square errors between both methods ranged from 0.036 to 0.055 s. High levels of agreement were observed for all variables.

Interpretation: These findings suggest that inertial measurement units can be used for outside laboratory assessment (e.g. in a hospital environment) of temporal gait parameters in the knee arthroplasty population.