Estimating running spatial and temporal parameters using an inertial sensor

An inertial measurement unit (IMU) is widely considered to be an economical alternative to capture human motion in daily activities. Use of an IMU for clinical study, rehabilitation, and in the design of orthoses and prostheses has increased tremendously. However, its use in defining running gait is limited. This study presents a practical method to estimate running spatial and temporal parameters using an inertial sensor by placing it on a shoe. A combination of a zero-crossing method and thresholding is used to identify foot-strike and foot-off based on foot acceleration during running. Stride time, ground contact time and flight time can then be identified. An off-phase segmentation algorithm is applied to estimate stride length and running speed. These two parameters are commonly used to evaluate running efficiency and to differentiate elite runners. This study found that an IMU can estimate foot-strike and foot-off with average absolute time differences of 2.60–6.04 and 2.61–16.28 ms, respectively. Stride time was estimated with error between − 4.04 and 0.33 ms. Stride length and running speed were estimated with maximum average errors of 45.97 mm and 0.41 km/h.