Computer modelling and experimental design of a gait orthosis for early rehabilitation of walking

Walking is a fundamental human activity [1]. Rehabilitation of walking is one of the essential goals for patients with spinal cord injury (SCI) or other neurological impairments [2, 3]. Early rehabilitation is desirable to maximise the beneficial effects, so training programmes should be initiated even when patients are still on bed rest. In order to promote early rehabilitation of patients with incomplete spinal cord injury who cannot maintain an upright posture, a Gait Orthosis for Early Rehabilitation (GOER) of walking was designed [2] and evaluated in this PhD work. This research started with a gait analysis experiment, through which the kinematics and kinetics of overground walking were investigated. Based on experimental walking data from able-bodied subjects, a least squares algorithm was developed to approximate the foot trajectories with circles. The determination of the best-fit circle for the toe trajectory over the whole gait cycle provided the basis for inducing toe movement by a rigid bar. Therefore a model of a two-bar mechanism was developed in Matlab/SimMechanics to simulate supine stepping. The simulated kinematics, including the angles of the hip, knee and ankle joints, showed comparable ranges of motion (ROMs) to the experimental walking performance in able-bodied subjects. This two-bar model provided the basis for the development of the GOER system. The intersegmental kinetics of the lower limb motion during supine stepping were investigated through computer simulation. A model of a leg linkage was firstly developed to simulate upright walking. After the model was validated by successful simulation of dynamic performance similar to experimental overground walking, the model was rotated by 90o to simulate stepping movement in a supine posture. It was found that the dynamics of the hip joint were significantly influenced by the position change from upright to supine, which highlighted the importance of a leg-weight support during supine stepping. In contrast, the kinetics of the ankle joint were much influenced by the forces applied on the foot sole which mimicked the ground reaction occurring during overground walking. Therefore a suitable force pattern was required on the foot sole in order to train the ankle joint during supine stepping. The simulated kinematic and kinetic results provided the basis for the design process of the GOER system. A GOER prototype with mechanisms for one leg was manufactured, which included a bar linkage to move the leg frame upwards and downwards and a cam-roller mechanism to rotate the shoe platform. The bar-cam GOER prototype achieved coordinated movements in the leg frame through constant rotation of an electric motor. Preliminary tests were carried out in three able-bodied subjects who followed the movements produced by the GOER prototype. The subjects felt walking-like stepping movement in the lower limb. Synchronised motion in the hip, knee and ankle joints was obtained, with the ROMs in the physiological ranges of motion during overground walking. The experimentally obtained joint profiles during supine stepping matched the simulated supine stepping and were close to the profiles during overground walking. Apart from inducing proprioceptive feedback from the lower limb joints, the GOER system required dynamic stimulation from the shoe platform to mimic load occurring during the stance phase of overground walking. Activated by pneumatic components, the shoe platform managed to apply forces on the foot sole with adjustable amplitudes. The pneumatic shoe platform was evaluated in ten able-bodied subjects and managed to induce walking-like pressure sensation on the foot sole with physiological responses from the leg muscles. In summary, this thesis developed and evaluated a new gait training robotic system targeting supine stepping for patients who are still restricted to a lying position. The conceptual design process was developed through computer modelling and it was implemented as a prototype. Evaluation tests on able-bodied subjects proved the technical feasibility of the robotic system for supine stepping and led to recommendations for further development