EEG driven model predictive position control of an artificial limb using neural net

Current research on neuroprosthetics is aimed at designing several computational models and techniques to trigger the neuro-motor rehabilitative aids. Researchers are taking keen interest to accurately classify the stimulated electroencephalography (EEG) signals to interpret the motor imagery/ execution. Some diseases or spinal cord injury completely destructs the sensory, motor and autonomous function for the limb movement. BCI (Brain computer Interface) provides a new communication pathway for those patients. Imagination of limb movements is used to operate a BCI. With analysis of acquired EEG signal due to motor imagery controlling of an artificial limb is possible. An Artificial Neural net (ANN) can be trained to predict the next position where the artificial limb moves to reach the target position. The neural net trained with desired position and current position of the arm as input and form that training it will be predict the next position. In this paper, a EEG driven model predictive position control of an artificial limb using neural net is presented. For this motor imagery EEG signal is acquired and the wavelet coefficients (feature) are extracted from the filtered EEG signal. The features are then classified. Here we classified left-right arm movement from the motor imagery by the backpropagation algorithm. The current position of the classified arm and the object position where the classified arm want to reach were fed to a neural net to predict the next position to move. The angles of shoulder and elbow joint were calculated from the next position by the inverse kinematics. By using State feedback PI controller the joints moves from the initial angle to the calculated angle to reach the target position. The procedure continues in a loop until the classified arm reaches to the target position.