EMG patterns induced in upper limb by haptic guidance for diagnosis and treatment evaluation

Robotic devices have been used successfully in assisted rehabilitation to increase, in continuity and intensity, the effects of neuromuscular stimulation. However, information resulting from the Patient-Robot interaction is not yet analyzed for diagnosis purposes. The neuromuscular reaction to a controlled stimulation, measured in terms of physiologic and mechanic variables, can be used to establish a common causal relation in healthy people to be used as a reference to evaluate the state of a patient, and thus, reinforce or reconsider its treatment. In this paper, we analyze the upper limb EMG patterns induced kinesthetically by uniform and cyclic trajectories using a haptic interface, and we show the existence of common patterns in healthy volunteers and how these patterns together with the mechanical response, can be used as a reference in diagnosis by means of a Similitude Index; we describe the EMG signals considered in this work, the experimental conditions to induce their activation patterns, and we discuss about some experimental results obtained with healthy volunteers and stroke patients.