Accelerometer-based step initiation control for gait-assist neuroprostheses

摘要

Electrical activation of paralyzed musculature cangenerate or augment joint movements required for walking aftercentral nervous system trauma. Proper timing of stimulation relativeto residual volitional control is critical to usefully affectingambulation. This study evaluates threedimensional accelerometersand customized algorithms to detect the intent to step fromvoluntary movements and trigger stimulation during walking inindividuals with significantly different etiologies, mobility limitations,manual dexterities, and walking aids. Three individualswith poststroke hemiplegia or partial spinal cord injury exhibitingvarying gait deficits were implanted with multichannel pulsegenerators to provide joint motions at the hip, knee, and ankle.An accelerometer integrated into the external control unit wasused to detect heel strike or walker movement, and wirelessaccelerometers were used to detect crutch strike. Algorithmswere developed for each sensor location to detect intent to stepto progress through individualized stimulation patterns. Testingthese algorithms produced detection accuracies of at least 90%on both level ground and uneven terrain. All participants usetheir accelerometertriggered implanted gait systems at homeand in the community; the validation/system testing was completedin the hospital. The results demonstrated that safe, reliable,and convenient accelerometerbased step initiation can beachieved regardless of specific gait deficits, manual dexterities,and walking aids.