Prosthesis Control Using a Nearest Neighbor Electromyographic Pattern Classifier

An investigation was conducted into the feasibility of applying a nearest neighbor algorithm to the problem of recognizing electromyographic (EMG) signal patterns for prosthesis control. A nearest neighbor algorithm correctly identified arm motions as belonging to one of six pattern classes from 72 to 100 percent of the time. A condensed nearest neighbor classifier was constructed to determine what minimum number of vectors was necessary in the look-up table.