A Method of Using Pressure Insoles for Foot Drop FES Gait Phase Detection

This paper reports on a approach to using a wireless plantar pressure insole to detect gait phase for foot drop functional electrical stimulation (FES). The approach normalizes the center of plantar (COP) and the peak pressure in four areas of the foot as input features of gait model, and simulates the pressure transfer during gait process, while avoiding poor sensor sensitivity of the sensor in certain regions under varus gait. This approach was incorporated into a fuzzy neural network (FNN) gait model which is used for real-time gait phase detection, and an artificial fish swarm algorithm (AFSA) developed to automatically acquire membership function parameters. A comprehensive experimentation was devised to test the performance of the gait model when dealing with simulated varus gait datasets from two different people, comprising a detailed analysis of results, which show our approach's successfully detected gait phase during varus gait.