A fuzzy logic gait event detector for fes paraplegic gait

Abstracf A system for automatic detection of the phases of gait of paraplegic subjects during Functional Electrical Stimulation (FES)-induced walking was developed. This fuzzy logic based detector is intended for on-line estimation of the patient's phase of gait, as part of a closed loop controller for improved paraplegic locomotion. The system was identified from sensor data using modified versions of fuzzy system identification algorithms. The best algorithm and fuzzy model configuration yielded a gait event detector that correctly predicted the patient's phase of gait 97% of the time using simulated noise free data and 94% of the time using experimental data. The modification of these algorithms reduced storage requirements by over 50% without a decrease in detection accuracy with noise free simulated data and with a decrease of 9% when noise ( m u . =lo% of input range) was added to all input data signals.