Validation of step time and asymmetry detection algorithm in high-intensity running with Movesense IMU

Movesense devices are open-source, low-cost, programmable with Bluetooth capability, offering applications from performance analysis to injury monitoring. Within a wider project aimed at developing comprehensive and freely-available assessment of human movement using wearable devices, this study aims to evaluate the validity and reliability of Movesense Inertial Measurement Units (IMUs) and novel algorithms for measuring step time and its right-left asymmetry during running. Data were collected using an app connected to the IMUs with custom firmware installed. The results were compared with those from a standard force plate system. Fourteen participants (7 females, 7 males; age: 22 ± 3 y; mass: 70.6 ± 8.7 kg) provided informed consent and performed a running protocol consisting of increments of 1 km/h every 2 min, starting at 8 km/h and continuing until exhaustion. Chest-mounted IMUs sampled at 208 Hz, while a force plate sampled at 1000 Hz. Step time was calculated by analysing vertical acceleration in both time and frequency domains, and vertical axis angular velocity was used to identify left and right foot strikes. The step time calculations resulted in a mean difference of 0 ± 6 ms (r = 0.97) across all 33,477 steps. All (100 %) left and right foot strikes were correctly identified. The step time imbalance (STI%) mean difference was 0.1 % ± 0.1 % (r = 0.92) for the 103 speed-by-participant subsets. The agreement is excellent between the Movesense IMU's utilising our custom algorithms and the criterion force plate system for step times, foot strike identification and step time asymmetry.