Integration of Smart Insoles for Gait Assessment in Exoskeleton-assisted Rehabilitation

A robotic exoskeleton enables individuals with limited or no mobility to engage in moderate exercises, promoting physical fitness and overall well-being. Exoskeletons, however, do not provide insights into gait patterns monitoring and analysis over time. This study proposes the integration of smart insoles as a cost-effective and non-invasive assistant for gait assessment in exoskeleton-assisted rehabilitation. The study, spanning 12 weeks, comprised three assessment sessions involving a total of 10 participants, including three healthy individuals, one stroke, one spinal cord injury, one traumatic brain injury, and four multiple sclerosis subjects. Gait phases were identified using a Finite State Machine with transitions guided by the predictions of a fuzzy c-means clustering algorithm. Kinematic and kinetic analyses highlighted disparities in stride time, cadence, stance time, and the trajectories of the centre of pressure. Despite these differences, individuals with multiple sclerosis showed no statistical differences from healthy subjects, underscoring the influence of the exoskeleton even in the absence of lower limb power. An analysis of acceptability demonstrated that participants found the smart insoles comfortable and expressed a willingness to employ them for future rehabilitation purposes. In conclusion, smart insoles can provide additional insights about users' gait compared to exoskeletons alone, allowing the progress of individuals involved in rehabilitation to be assessed over time, and helping clinicians to develop tailored rehabilitation plans.