Neural Control of the Computer Cursor Based on Spectral Analysis of the Electromyogram

A classification algorithm is developed to translate facial movements into five cursor actions: i) left, ii) right, iii) up, iv) down, and v) left-click. The algorithm utilizes the unique spectral characteristics exhibited by electromyogram (EMG) signals obtained from different muscles in the face to assist in the classification process. A previous three-electrode, EMG-based system was utilized in to perform a similar translation of facial movements into cursor actions. This system also made use of spectral analysis to classify muscle activity. It was found that this system does not always discriminate between the EMG activity assigned to up and down cursor actions efficiently. To remedy this matter, a fourth electrode was added and a new classification algorithm was devised. This paper details the classification algorithm utilized with the four-electrode system. It also compares the effectiveness of the four-electrode system to that of the three-electrode system in classifying EMG activity into cursor actions. This was done through the use of Matlab simulations. It will be shown that the new four-electrode system produces significant improvements in classification performance