Analysis of the variability of gait cycle parameters based on upper body acceleration data obtained from Inertial sensors to predict the risk of falling

Variability in the values of gait cycle parameters is a measure of walking stability. Mental or physical disabilities usually manifest themselves in hand or foot movements. In this study, using the upper body acceleration data which was obtained from Inertial sensors, the variability of gait cycle parameters was investigated. Then, the reproducibility of the parameters in healthy individuals in this method was compared with the gold standard methods for healthy and sick elderly (at risk). The results showed that the step time parameter could be a suitable criterion as a standard method for separating healthy and sick people (at risk) as it had a good correlation with gold standard methods. Also, the variability of the step time parameter with a slight error compared to the gold standard methods can be considered as a suitable criterion; however, the variability of the swing time parameter showed a significant difference compared to the conventional methods. Investigating the variability of acceleration size in three anatomical directions showed that the acceleration in the anterior-posterior and lateral directions had the highest repetition among healthy subjects.