Patient satisfaction with body-powered myoelectric upper limb prostheses is limited, often resulting in device abandonment. Multifunctional hand prostheses are one potential solution to increase patient acceptance. These require sophisticated control schemes like patternrecognition- based approaches involving classification of myoelectric signals (MES). To allow fast and flexible evaluation of prosthesis control approaches, a prototyping environment based on the Raspberry Pi and MATLAB/Simulink was created. It supports commonly applied features like RMS and zero crossings as well as classification methods like Naive Bayesian and Support Vector Machine classifiers. After classifier training with a custom MATLAB application, MES can be classified in real-time and the results employed for prosthesis actuation. The setup was tested with five participants for controlling a Michelangelo Hand. Over 90% of movements were correctly identified for three classes from two channel EMG data.
CITATION STYLE
Attenberger, A., & Buchenrieder, K. (2015). MATLAB/simulink-supported EMG classification on the raspberry pi. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9520, pp. 449–456). Springer Verlag. https://doi.org/10.1007/978-3-319-27340-2_56
Mendeley helps you to discover research relevant for your work.