Characterization of lower limb activity during gait using wearable, multi-channel surface EMG and IMU sensors

Lower limb amputees who wear a conventional passive prosthesis usually consume more metabolic energy than normal persons and may put more pressure on their knees, which eventually results in knee damage. In order to model the muscle activity during human gait, a simulated amputee walking on normal subjects was investigating by adding a 10 lbs. load on their lower right limb during level walking and walking on a slope of 10 degrees. Multi-channel surface electromyogram (sEMG) signals from 8 muscle groups were recorded and modeled. Additionally, acceleration on different types of walking was recorded from three positions on the lower body. Several basic features extracted from EMG data were explored. Results shows that the human gait associates with a complex collaboration between various muscles and the effect of load on gait was clearly revealed by the EMG signals recorded from the lower limb muscles. This study was implemented to provide an initial assessment for the optimization of active biomechanical lower limb prostheses.