Abstract
Balance assessment is a key metric for tracking the health and fall risk of individuals with balance impairment. Leveraging wearable sensors and mobile devices can increase clinical accessibility to objective balance metrics. Previous work has been conducted validating center of mass (COM) acceleration metrics from mobile devices against the gold standard force plate center of pressure (COP) position; however, most studies have been restricted to devices being placed close to the subject's COM. In this study, rigid body kinematics and the inverted pendulum model were used to develop a novel methodology for calculating COM acceleration using mobile devices in arbitrary positions, as well as an approach for conversion of COM measurements to COP position for direct validation with force plate measurements. Validation of this methodology included a direct comparison of smartphone and force plate results for COM accelerations and COP positions, as well as statistical comparisons using Spearman's rank correlation. The results show strong analysis performance for both approaches during a subject's intentional swaying, with more limited results in cases of little motion. The strong performance warrants future work to further improve accessibility by removing dependence on motion capture systems or replacing them with cost-effective alternatives. The accurate tracking of COM acceleration and COP position information for mobile devices at arbitrary positions increases the flexibility for future mobile or at-home balance assessments.