Development and Electromyographic Validation of a Compliant Human-Robot Interaction Controller for Cooperative and Personalized Neurorehabilitation
Stefano Dalla Gasperina,Valeria Longatelli,Francesco Braghin,Alessandra Pedrocchi,Marta Gandolla
DOI: https://doi.org/10.3389/fnbot.2021.734130
IF: 3.493
2022-01-18
Frontiers in Neurorobotics
Abstract:Background Appropriate training modalities for post-stroke upper-limb rehabilitation are key features for effective recovery after the acute event. This study presents a cooperative control framework that promotes compliant motion and implements a variety of high-level rehabilitation modalities with a unified low-level explicit impedance control law. The core idea is that we can change the haptic behavior perceived by a human when interacting with the rehabilitation robot by tuning three impedance control parameters. Methods The presented control law is based on an impedance controller with direct torque measurement, provided with positive-feedback compensation terms for disturbances rejection and gravity compensation. We developed an elbow flexion-extension experimental setup as a platform to validate the performance of the proposed controller to promote the desired high-level behavior. The controller was first characterized through experimental trials regarding joint transparency, torque, and impedance tracking accuracy. Then, to validate if the controller could effectively render different physical human-robot interaction according to the selected rehabilitation modalities, we conducted tests on 14 healthy volunteers and measured their muscular voluntary effort through surface electromyography (sEMG). The experiments consisted of one degree-of-freedom elbow flexion/extension movements, executed under six high-level modalities, characterized by different levels of (i) corrective assistance, (ii) weight counterbalance assistance, and (iii) resistance. Results The unified controller demonstrated suitability to promote good transparency and render both compliant and stiff behavior at the joint. We demonstrated through electromyographic monitoring that a proper combination of stiffness, damping, and weight assistance could induce different user participation levels, render different physical human-robot interaction, and potentially promote different rehabilitation training modalities. Conclusion We proved that the proposed control framework could render a wide variety of physical human-robot interaction, helping the user to accomplish the task while exploiting physiological muscular activation patterns. The reported results confirmed that the control scheme could induce different levels of the subject's participation, potentially applicable to the clinical practice to adapt the rehabilitation treatment to the subject's progress. Further investigation is needed to validate the presented approach to neurological patients.
robotics,computer science, artificial intelligence,neurosciences