A Reliability Study of Small, Portable, Easy-to-Use, and IMU-Based Sensors for Gait Assessment

The standard motion analysis systems are limited to laboratory settings; therefore, an individual's gait may not be realistic, as it is removed from the day-to-day environment in which people ambulate. The modern world and advanced technologies have driven portable, affordable, and wearable sensors for real-world gait assessment that can be used outside the laboratory and during day-to-day activities. Wearable sensors offer a promising solution; however, despite that, the reliability of many wearable systems, especially under unsupervised and real-world-like conditions, remains insufficiently validated. This study aimed to establish intra- and inter-rater reliability of the inertial sensors as a tool used in gait analysis in a quasi-real-world environment. Ninety-eight healthy participants (52% females) aged 19-33 years took part in this reliability study. The research procedures included two separate measurements of gait analysis at participants' preferred walking speed, conducted by two raters assessing selected spatiotemporal parameters. The reliability was calculated using intraclass correlation coefficients (ICC), and the bias was assessed using the Bland-Altman method. The analysis of intraclass correlation coefficients (ICC) revealed excellent, or near-excellent, reliability for walking speed, cadence, and stride length between raters (ICC = 0.932-0.941, 0.950-0.957, and 0.916-0.938, respectively) and between measurements (ICC = 0.916-0.928, 0.934-0.938, and 0.888-0.906, respectively). Bland-Altman plots confirmed minimal systematic bias for both inter- and intra-rater assessments, with differences in walking speed below 0.038 km/h, cadence below 0.283 steps/min, and stride length below 0.827 cm. The examined sensors are reliable tools for walking speed, cadence, and stride length in a quasi-real-world environment gait assessment. Future studies should include gait analysis involving random path and direction changes, turns, uneven or slippery surfaces, and natural environments. Additionally, research should consider individuals ambulating with various walking aids, or those with unilateral disorders, such as stroke.