Biomechanical Effects of Providing Assistance with an Ankle Exoskeleton for People with Hemiplegic Gait

Hemiplegia is one of the main causes of disability for adults in the European Union and in the world. About 80 % of affected people experience deficits in motor control, which cause problems in gait and balance. There are some assistive devices that help people with hemiplegia during gait but they have some adverse effects that reduce the improvement of rehabilitation outcome. Nowadays, ankle exoskeletons are being developed to provide gait assistance and improve the rehabilitation process. ABLE-S is a wearable, unilateral, powered ankle exoskeleton that provides adaptive assistance to improve forward propulsion and foot clearance in people with hemiplegia. During the de- velopment of this thesis, an experimental protocol has been designed in order to: (1) test the feasibility of the exoskeleton ABLE-S, (2) collected the subjective opinion of the participants by questionnaires and (3) analyze the effects of control parameters on clinically meaningful biomechanical metrics to further develop adaptive controllers. Tests with participants with chronic hemiplegia (n = 9) showed that walking with ABLE-S cor- rected foot drop significantly (25 % of improvement) while reducing hip compensatory move- ments by 16 %. Asymmetrical temporal and spatial gait patterns were reduced (17 % and 39 %, respectively) leading to less exhaustion together with better balance control. Finally, the device also increased the foot tilting angle by 378 %. 78 % of the participants indicated that they were satisfied with their performance walking with the robotic assistance and 56 % would use the device for community ambulation. The assistance and weight of the device were the features that received the best and worst scores respectively. Relationships for changes in both DF and PF peak magnitudes, and PF peak timing were ex- tracted. Regarding DF peak magnitude, a significant relation was only found for the temporal symmetry index. Changes in the applied PF parameters were not only correlated with a variety of temporal and spatial parameters but also with compensatory movements, e.g., hip hiking. The high influence of the robotic plantarflexion action compared to the dorsiflexion one might highlight that control strategies for hemiplegia rehabilitation should focus on providing adap- tive plantarflexion assistance based on the level of impairment. The obtained results from the tests performed showed that ABLE-S assistance can improve the gait pattern of individuals with mild-to-moderate gait impairments from hemiplegia as par- ticipants received relevant benefits after only 15 minutes of practice. The results of these tests highlight the need of design iterations focusing on weight and adaptability to increase the per- formance, usability and acceptability of the device. The results also helped to propose an adap- tive control strategy to adapt the three control parameter of the exoskeleton based on real-time data capture from wearable sensors presented in ABLE-S. Future work should include larger cohorts with higher acclimation periods and training time to explore the upper limit of ankle exoskeleton assistance. Studies have to be composed of multiple training sessions to study motor learning and time to adapt to the exoskeleton