Design and Development of Assistive Ankle Exoskeleton for Rehabilitation using Electromyography

Powered ankle exoskeletons can be used in gait rehabilitation and gait assistance. For this purpose, the design must be light weight, comfortable and controlled by the user. In this study, a two degree of freedom light weight ankle exoskeleton (0.5 Kg at ankle, 0.2 Kg at waist) was developed with an EMG driven Brain Computer Interface based control to provide powered mechanical assistance according to motion intention of the user. The developed brain computer interface uses EMG pattern recognition with previously tested signal processing and feature extraction techniques with only a small window time of 100 milliseconds. The design was tested on 08 healthy subjects with high offline $(98.4 \pm 1.0$%) and online accuracies ($92.1\pm 2.0$%) accuracies ($p\lt0.05)$. The design also incorporates force measurements at sole for gait phase recognition.