Estimation of gait parameters in healthy and hemiplegic individuals using Azure Kinect: a comparative study with the optoelectronic system
Serena Cerfoglio, Claudia Ferraris, Luca Vismara, Gianluca Amprimo, Lorenzo Priano, Matteo Bigoni, Manuela Galli, Alessandro Mauro, Verônica Cimolin
Abstract
Introduction: Walking ability is essential for maintaining functional independence, but it can be impaired by conditions like hemiplegia resulting from a stroke event. In post-stroke populations, accurately assessing gait anomalies is crucial for rehabilitation to promote functional recovery, and to prevent falls or injuries. Methods: The aim of this study is to evaluate gait-related parameters using a solution based on a single RGB-D camera, specifically Microsoft Azure Kinect DK (MAK), on a short walkway in both healthy (n= 27) and post-stroke individuals with hemiplegia (n= 20). The spatio-temporal and center of mass (CoM) parameters estimated by this approach were compared with those obtained from a gold standard motion capture (MoCap) system for instrumented 3D gait analysis. Results: The overall findings demonstrated high levels of accuracy (> 93%), and strong correlations (r > 0.9) between the parameters estimated by the two systems for both healthy and hemiplegic gait. In particular, some spatio-temporal parameters showed excellent agreement in both groups, while CoM displacements exhibited slightly lower correlation values in healthy individuals. Discussion: The results of the study suggest that a solution based on a single optical sensor could serve as an effective intermediate tool for gait analysis, not only in clinical settings or controlled environments but also in those contexts where gold standard systems are not feasible.