Title :
Inducing self-selected human engagement in robotic locomotion training
Author :
Collins, Steven H. ; Jackson, Robert W.
Author_Institution :
Dept. of Mech. Eng., Carnegie Mellon Univ., Pittsburgh, PA, USA
Abstract :
Stroke leads to severe mobility impairments for millions of individuals each year. Functional outcomes can be improved through manual treadmill therapy, but high costs limit patient exposure and, thereby, outcomes. Robotic gait training could increase the viable duration and frequency of training sessions, but robotic approaches employed thus far have been less effective than manual therapy. These shortcomings may relate to subconscious energy-minimizing drives, which might cause patients to engage less actively in therapy when provided with corrective robotic assistance. We have devised a new method for gait rehabilitation that harnesses, rather than fights, least-effort tendencies. Therapeutic goals, such as increased use of the paretic limb, are made easier than the patient´s nominal gait through selective assistance from a robotic platform. We performed a pilot test on a healthy subject (N = 1) in which altered self-selected stride length was induced using a tethered robotic ankle-foot orthosis. The subject first walked on a treadmill while wearing the orthosis with and without assistance at unaltered and voluntarily altered stride length. Voluntarily increasing stride length by 5% increased metabolic energy cost by 4%. Robotic assistance decreased energy cost at both unaltered and voluntarily increased stride lengths, by 6% and 8% respectively. We then performed a test in which the robotic system continually monitored stride length and provided more assistance if the subject´s stride length approached a target increase. This adaptive assistance protocol caused the subject to slowly adjust their gait patterns towards the target, leading to a 4% increase in stride length. Metabolic energy consumption was simultaneously reduced by 5%. These results suggest that selective-assistance protocols based on targets relevant to rehabilitation might lead patients to self-select desirable gait patterns during robotic gait training sessions, possibly facilitating better a- herence and outcomes.
Keywords :
gait analysis; medical robotics; mobile robots; patient treatment; energy-minimizing drives; gait patterns; manual treadmill therapy; metabolic energy consumption; mobility impairments; paretic limb; patient exposure; robotic assistance; robotic gait training; robotic locomotion training; selective-assistance protocols; self-selected human engagement; Legged locomotion; Medical treatment; Muscles; Robot kinematics; Torque; Training;
Conference_Titel :
Rehabilitation Robotics (ICORR), 2013 IEEE International Conference on
Conference_Location :
Seattle, WA
Print_ISBN :
978-1-4673-6022-7
DOI :
10.1109/ICORR.2013.6650488