An intelligent strategy for motion interpretation in an EMG-controlled prosthesis

A heuristic rule-based strategy is proposed for the interpretation of motion patterns in an electromyogram (EMG)-controlled upper extremity prosthesis. The philosophy of the approach is to overcome the ambiguity in motion interpretation using a number of noise interpretation rules. Results of computer simulations on an above-elbow amputee show a high classification performance for a two-degree-of-freedom prosthesis.<>