Using 3D skeleton tracking for gait analysis in older adults compared to a kinect-based mobility analysis system

Background Older persons are affected by falls and their associated consequences. This cohort is often unaware of their own risk of falling or misjudge it. Mobility and fall risk apps are therefore useful tools for recognizing fall risks and creating opportunities to reduce them. The aim of the study was to compare the agreement of a mobility analysis smartphone camera-based application (SCA) with an included advanced 3D skeleton tracking with the reference system Microsoft Kinect System R (MKS). Methods In a cross-sectional design, 186 data sets from 31 older adults were assessed at the same time with both systems depending on two gait modes (comfort and fast). Results For the comfort mode, the ICC values ranged from 0.74 (gait speed; 95% CI: 0, 0.93) to 0.85 (cadence; 95% CI: 0.62, 0.93). The fast mode showed a slightly lower agreement between both systems. ICC values moved from 0.66 (gait speed; 95% CI: 0, 0.90) to 0.81 (cadence; 95% CI: 0.46, 0.91). Conclusion With the exception of gait speed, both systems demonstrated good agreement (ICC > 0.75) in measuring spatiotemporal gait parameters. These findings suggest that the SCA has the potential to serve as a reliable and practical instrument in the context of clinical gait analysis.