Comparative analysis of machine learning algorithms on myoelectric signal from intact and transradial amputated limbs

Abstrak

Control strategies of smart hand prosthesis-based myoelectric signals inrecent years don't provide the patients with the sensation of biologicalcontrol of prostheses hand fingers. Therefore, in current workhyperparameters optimization in machine learning algorithm and handgesture recognition techniques were applied to the myoelectric signal-basedon residual muscles contraction of the amputees corresponding to intactforearm limb movement to improve their biological control. In this paper,myoelectric signals are extracted using the MYO armband to recognize tengestures from ten volunteers (healthy and transradial amputation) on theforearm, thereafter the noise of myoelectric signals using a notch filter (NF)is removed. The proposed classification system involved two machinelearning algorithms: (1) the decision tree (DT), tri-layered neural network(TLNN), k-nearest-neighbor (KNN), support vector machine (SVM) andensemble boosted tree (EBT) classifiers. (2) the optimized machine learningclassifiers, i.e., OKNN, OSVM, OEBT with optical diffraction tomography(ODT) and ommatidia detecting algorithm (ODA). The experimental resultsof classifiers comparison pointed out an algorithm that outperformed withhigh accuracy is OEBT closely followed by OKNN achieves an accuracy of97.8% and 97.1% for intact forearm limb, while for transradial amputationwith an accuracy of 91.9% and 91.4%, respectively.