Traditional gait analyses require the use of purpose-built laboratories and specialized equipment. While these enable full quantification of gait mechanics including ground reaction forces, joint kinematics, joint moments, and spatiotemporal parameters, they also limit the both the types of gait that can be analyzed (often only level, steady-state gait) and the amount of data that can be analyzed. Wearable inertial sensors, which can be used outside of laboratories, enable the measurement of human movement in real-world settings. However, making use of real-world data from wearable inertial sensors is challenging for two primary reasons: 1) wearable sensors measure linear acceleration and angular velocity, which is fundamentally different than the traditional gait lab measures of force, position, and orientation and 2) many types of walking (e.g., short walks, stairs, long walks, slopes, etc.) occur in the real world and therefore gait must be characterized before analysis (e.g., how do we identify ‘steady-state’ gait?). Our research is focused on advancing the use of inertial sensors for real-world gait analyses, which we believe will be instrumental in facilitating new research discoveries and health interventions.
Overhead view of sample out-of-lab gait data. The walking path (foot trajectory) calculated from foot sensor data is illustrated by the colored line. Steady-state strides are colored red/orange/yellow according to stride speed and non-steady-state strides are blue.