Designing Sonified Feedback on Knee Kinematics in Hemiparetic Gait Based on Inertial Sensor Data

In recent years, interactive sonification based on data from wearable sensors has been explored as a feedback tool in movement rehabilitation. However, it is yet to be routinely adopted as part of physiotherapy protocols, partly due to challenges with designing solutions tailored to diverse patients. In this work, we propose a set of adaptable feedback paradigms on knee kinematics for hemiparetic stroke patients undergoing gait training. We first collected inertial data and video footage from 15 hemiparetic patients during overground walking. The video footage was then analyzed by a physiotherapist, who identified three main knee-related movement impairments - reduced range of motion, dysregulated extension, and hyperextension. Using a custom-built software architecture, we devised two music-based paradigms for providing tailored concurrent feedback on knee movement and the impairments identified by the physiotherapist
based on inertial data. The paradigms will be clinically tested with patients as part of a future study, and we believe that their impairment-specificity and individual adjustability will make them an advancement of existing auditory feedback designs.