Controlling the hand and forearm movements of an orthotic arm using surface EMG signals

For people affected by paralysis, a support system for moving the non-functional body parts is highly essential. Automated orthotic arm is one such device. In the proposed work the authors have replicated the movements of the healthy arm on to an orthotic arm using surface electromyography (EMG) signals. The main objective of the paper is to first recognize the hand and forearm movements and later identify the three static positions, like relax, semi-flex and flex. For this purpose certain statistical features reflecting the movements are extracted. The support vector machine (SVM) classifier used resulted in 86.325% accuracy, in classifying the hand and forearm movements. For the three positions of both the hand and forearm an accuracy of 100% is recorded.