Reduced Daily Re-calibration of Myoelectric Prosthesis Classifiers Based on Domain Adaptation

Control scheme design based on surface electromyography (sEMG) pattern recognition has been the focus of much research on myoelectric prosthesis (MP) technology. Due to inherent non-stationarity in sEMG signals, prosthesis systems may need to be re-calibrated day after day in daily use applications, thereby hindering MP usability. In order to reduce the recalibration time in the subsequent days following the initial training, we propose a domain adaptation (DA) framework, which automatically reuses the models trained in earlier days as input for two baseline classifiers: a polynomial classifier (PC) and a linear discriminant analysis (LDA). Two novel algorithms of DA are introduced, one for PC and the other one for LDA. Five intact-limbed subjects and two transradial-amputee subjects participated in an experiment lasting 10 days, to simulate the application of a MP over multiple days. The experiment results of four methods were compared: PC-DA (PC with domain adaptation), PC-BL (baseline PC), LDA-DA (LDA with domain adaptation), and LDA-BL (baseline LDA). In a new day, the DA methods reuse nine pre-trained models which were calibrated by 40 seconds training data per class in nine previous days. We show that the proposed DA methods significantly outperform non-adaptive baseline methods. The improvement in classification accuracy ranges from 5.49% to 28.48%, when the recording time per class is 2 seconds. For example, the average classification rates of PC-BL and PC-DA are 83.70% and 92.99% respectively for intact-limbed subjects with a 9-motions classification task. These results indicate that DA has the potential to improve the usability of MPs based on pattern recognition, by reducing the calibration time.