Task estimation of upper-limb using EEG and EMG signals

Many kinds of wearable robots have been proposed. In the control of those robots, surface electromyogram (sEMG) signals are widely used in order to estimate a user's motion intention. However, EMG signals that are needed to estimate a user's motion are not always available with all users. On the other hand, in recent years, an electroencephalogram (EEG) signal that is measured through an electrode on the scalp is used to control robots. It is not easy to estimate a user's motion-intention from measured EEG signals in comparison with sEMG signals in which the increase or decrease of the signals relates closely to the motion. In this study, an electric artificial arm for above elbow amputees is controlled based on EMG and EEG signals. An EEG-based control method is proposed to control the forearm and wrist motions of an electric artificial arm in this paper. The target position of the hand is estimated based on the EEG signals, the shoulder and elbow motions. In the proposed method, the target position is selected based on the shoulder and elbow motions, then EEG signals are used to judge whether the selected target position based on the shoulder and elbow motions is correct.