Reliability and Validity of a Wearable Inertial Sensor System for Gait Assessment in Healthy Young Adults

Abstract

Wearable inertial sensors are considered to be low cost, portable and user-friendly for gait assessment. But relatively little is known about their reliability and validity compared with the other well-established, sophisticated techniques. This study aimed to determine the reliability and validity of an inertial sensor system (APDM Mobility Lab). A total of 20 participants were commanded to take a 10-m walk at their self-selected speeds. A series of gait parameters were collected and analyzed by this wearable gait analysis system and a three-dimension (3-D) motion analysis system simultaneously. The reliability of APDM system was evaluated using interclass correlation coefficient (ICC). The Pearson correlation analysis was used to evaluate the relations of spatiotemporal parameters from the two systems. Furthermore, Bland-Altman analysis was applied to evaluate the validity of APDM system. Results showed that the inertial sensor system of APDM Mobility Lab demonstrated excellent reliability $(\text{ICC}=0.905-0.991)$. The parameters analyzed by APDM system and 3-D motion capture system showed moderate to high correlations for stride length, mean velocity and cadence $(\mathrm{r}= 0.551-0.875)$. Moreover, Bland-Altman analysis demonstrated that the bias of all parameters approached to zero, particularly for the stride time, stride length and mean velocity. The wearable inertial sensor-based system used in this study is a comparatively reliable and valid tool for spatiotemporal gait assessment in healthy adults.