FES Gait Re-education: The Swing Phase Estimation

This paper presents the use of multiple sensors for walking assessment and provision of cognitive feedback during early re-education of incomplete spinal cord injured (SCI) humans. The paper is focused on the swing phase estimation as an important part of the Functional Electrical Stimulation (FES) gait re-education system for incomplete spinal cord injured persons. The proposed sensory system comprises four accelerometers, one gyro placed at the shank of the paretic leg, and two goniometers placed at the knee and ankle joints, respectively. The data from the sensors are input in the mathematical algorithm applied for swing quality estimation. The output from the algorithm is a numerical value. The calculated output is divided into three levels, each defining the swing quality in terms of good, sufficient, and poor. This information is provided to the patient as an auditory signal. The patient is taught to maximize his efforts to improve the quality of walking, that is, to move the more affected leg in a way that will generate the auditory output corresponding to the level "good".  The preliminary measurements were performed in healthy subjects walking on even terrain and in an incomplete SCI person with C6 lesion during walking on the treadmill. FES in the latter case was triggered manually by a physiotherapist. The results showed that the timing of FES triggering played an important role in sensory-supported FES-assisted walking, that is, the auditory feedback was also a cue to the therapist controlling the FES. The swing quality estimation enabled patients to voluntarily improve their walking, consequently the intensity of FES assistance was decreased. This suggests that the use of an FES multisensor system for cognitive feedback is efficient rehabilitative method in early stage of rehabilitation of walking.