Neural network-based gait assessment using measurements of a wearable sensor system

A wearable gait analysis system has been developed for ambulatory measurement of gait in long-term experiments and daily life applications. Based on the measurement results of the developed system in a dynamic validation experiment, we trained neural networks for the estimation of joint angle, joint force, and joint moment using the ground reaction forces (GRFs) and moments in the gait analysis. These kinetic and kinematic parameters can be estimated through the neural networks trained especially for one person, but the possibility of building a universal model for most people should be studied in further research.