Foot Drop Correcting System Design Using Functional Electrical Stimulation

This thesis introduces new strategies to improve the adaptivity and quality of FES for foot drop correction. Based on these strategies, a self-adaptive foot drop stimulator has also been developed to demonstrate the improvement over current systems. Methodology:The electromyogram (EMG) of the tibialis anterior muscles of healthy subjects was recorded during walking. These EMG data were then normalized and averaged to serve as the dataset of stimulation envelopes. To make the stimulation more adaptive, the gaits of another group of subjects were also recorded for testing several algorithms of predicting stride frequency from previous strides. The best algorithm was then selected to be used in the new system, which pre-stores the envelope dataset for outputting proper stimulation patterns. The clinical evaluation of the system was conducted on a post-stroke foot drop patient. Results:The clinical test shows that the newly developed system is significantly better than a system using the configurations that are common in commercial systems. The improvement might be due to the incorporated EMG dataset for modulating stimulations, as well as the carefully selected algorithm for stride frequency prediction, which outperforms the widely-used algorithm on the stride dataset.The main creativities of this work are:1. Using healthy TA EMG dataset sampled during normal walking overground modulate FES output envelop;2. Using new stride frequency estimation method to select output envelop from the TA EMG dataset;3. Using the TA EMG dataset and the new stride frequency estimation method build a self-adaptive one channel functional electrical stimulation based foot-drop corrector, and have a pilot test experiment on one post-stroke foot-drop patient.