Fig. 2

(A): The closed loop prosthesis control scheme. The linear envelope of EMG recorded by the electrodes was filtered and normalized, and then translated to prosthesis commands (continuous signal) and feedback levels (discretized). The tactors were activated to indicate the level of the myoelectric signal amplitude and each level was associated with a specific spatial stimulation pattern. The participants could use vision to monitor the movement of the prosthesis and closing of the pin. In the condition with no supplementary feedback, the EMG feedback was deactivated (switch). (B): The steps of the experimental task. The trial started with the participants generating a muscle activation which initiated the prosthesis movement. The task for the subjects was to grasp a pin and produce a force that was just enough to open it. Once the prosthesis grabbed a pin, a low force would open it only halfway. The correct force would open the pin fully, with the pin handles just touching. Further increase of the force would result in a failed trial