Upper limb prosthesis control through integration of EMG signals and compensatory body motions

In recent years, the development of advanced technologies for upper limb prostheses has led to numerous technological improvements aimed at making prosthesis use as natural and intuitive as possible. However, a major challenge remains in the control system, which currently relies solely on interpreting the myoelectric signals from the residual limb muscles. The limitations of this approach often lead patients to abandon the active component of the prosthesis, instead using it as a passive rigid structure and performing tasks through compensatory movements of the torso and shoulder. To address this issue, this thesis proposes the use of these compensatory movements as error signals to enhance prosthetic control. This approach has been specifically applied to a wrist with one degree of freedom, where compensatory movements were detected using an IMU sensor. In addition, an architecture was developed to integrate EMG and IMU signals, enabling simultaneous use of the wrist and hand. Finally, experiments were conducted with able-bodied and amputee participants to validate the proposed control strategy.