Validation of an inertial measurement unit for the quantification of rearfoot kinematics during running

Abstract

BackgroundThe popular protocol used to study running motion suffers from problems that lead to a limited ability to generalize the obtained results. Inertial measurement units (IMU) appear to be promising in increasing ecological validity of the collected data. However, quantifying running kinematics utilizing IMU signals is complex and potentially affected by several well-established and less well-known errors. Research questionThe purpose of this study was to examine the validity of kinematic variables obtained from a single, shoe-mounted IMU using an opto-electronic motion analysis reference system. Methods51 recreational runners were analyzed, performing a single continuous run at three different speeds (10, 12, 15 km/h) on a treadmill. Descriptive statistics (Bland & Altman analysis, box plots, scatter plots) were employed to analyze the agreement between the two instruments. ResultsThe findings of this study revealed considerable systematic and large random disagreement, which, in turn, is characterized by substantial inter-individual differences in the error distribution. These discrepancies may partly be explained by differences in foot strike behavior, resulting in varying degrees of vibration impact acting on the IMU. SignificanceAdvances in IMU technology, as well as exploring new application approaches and signal processing strategies, might enhance the usability of IMUs in analyzing running kinematics.