Design and development of a tool to facilitate learning to walk with a powered prosthesis

Abstract

The current paper presents research on a segment of gait analysis and studies to what extent prostheses' designs are able to mimic the behaviour of the lost limb while blending into the locomotion system. This is achieved by producing a tool and associated methodology that facilitates the process of learning to walk with a powered prosthesis. Simulation results of real time event detection at ground walking at subjects' preferred pace yielded a detection accuracy of 95.0%. Obtaining the time spent on monitored limb as proportion of the gait cycle allows correction of one of the principal sources of gait asymmetry in amputees, with the help of an audio feedback tool. This opens up the potential to construct simple and efficient instrument that targets gait asymmetry. A shift to a more suitable environment for real time gait processing should be considered before enhancing and increasing the complexity of the system.