Over three decades of upper-limb robotic neurorehabilitation: drawing conclusions and future work

Robotic rehabilitation devices for upper limb impairment emerged in the 1990s, notably with the development of the MIT Manus. Early studies suggested that robotic manipulation could enhance post-stroke recovery and introduce innovative strategies for neurologic rehabilitation. The proliferation of robotic devices in the 2000s yielded promising results, demonstrating the efficacy of upper-limb robotic training in improving outcomes for chronic stroke patients. These studies showed that robotic interventions could attenuate chronic neurologic deficits beyond the expected recovery period post-stroke. Moreover, research revealed that the neuro-recovery process extends well beyond the commonly accepted three-month post-stroke interval, with sustained improvements observed even after three years. However, the findings of a large multicenter study, RATULS, conducted in 2019, cast doubt on the routine use of robotics in rehabilitation. This study questioned the effectiveness of robotics, citing a lack of additional benefits compared to standard care and raising concerns about cost-effectiveness. These findings prompt reflection on the necessity for devices capable of delivering substantially superior functional outcomes. We believe it is crucial to revisit the rationale for employing robotics in upper-limb rehabilitation, alongside the principles of motor relearning, to address issues such as patient selection through instrumental evaluation and the formulation of new personalized intervention paradigms. In this paper, we discuss these matters, which have led us to propose an intervention based on 3D robot-assisted functional movements, briefly outlined herein.