You are on page 1of 5

Electromyogram pattern

recognition for control of


powered upper-limb prostheses:
State of the art and challenges
for clinical use
Erik Scheme, MSc, PEng; Kevin Englehart, PhD, PEng

This article and any supplementary material should be cited as follows: Scheme E, Englehart K. Electromyogram pattern recognition
for control of powered upper-limb prostheses: State of the art and challenges for clinical use. J Rehabil Res Dev. 2011;48(6):643–60.
DOI:10.1682/JRRD.2010.09.0177
• Study Aims
– Describe issues and best practices in electromyogram
(EMG) pattern recognition.
– Identify major challenges in deploying robust control.
– Advocate research directions.
• Relevance
– Using EMG signals to control upper-limb prostheses offers
autonomy of control via residual muscle contraction.
– Pattern recognition to discriminate multiple degrees of
freedom has shown great promise in research literature
but not yet a clinically viable option.

This article and any supplementary material should be cited as follows: Scheme E, Englehart K. Electromyogram pattern recognition
for control of powered upper-limb prostheses: State of the art and challenges for clinical use. J Rehabil Res Dev. 2011;48(6):643–60.
DOI:10.1682/JRRD.2010.09.0177
• Simple one-muscle one-function approach to
conventional control is naïve to complexities
of EMG cross talk, muscle co-activation, and
contribution of deep muscle.
• This has motivated use of pattern-recognition
approach to myoelectric control.
– By using multiple EMG sites, effective feature
extraction, and multidimensional classifiers, one
can achieve control of many more classes of
motions.
This article and any supplementary material should be cited as follows: Scheme E, Englehart K. Electromyogram pattern recognition
for control of powered upper-limb prostheses: State of the art and challenges for clinical use. J Rehabil Res Dev. 2011;48(6):643–60.
DOI:10.1682/JRRD.2010.09.0177
Pattern Recognition

Stages of signal processing for EMG pattern recognition. All


approaches to EMG pattern recognition go through these
fundamental processing stages. Feature-extraction stage
increases information density of EMG signals.

This article and any supplementary material should be cited as follows: Scheme E, Englehart K. Electromyogram pattern recognition
for control of powered upper-limb prostheses: State of the art and challenges for clinical use. J Rehabil Res Dev. 2011;48(6):643–60.
DOI:10.1682/JRRD.2010.09.0177
Conclusions
• Best Practices
– For slowly varying EMG patterns, time domain features offer suitable
tradeoff in accuracy and computational complexity.
– With appropriate feature set and sufficient channels, most modern
classifiers will perform similarly. However, linear discriminant analysis
is easy to implement and train.
– Most meaningful assessment is function user derives from device.
• Major Challenges
– Electrode shift, variation in force, variation in position of limb, and transient
changes in EMG.
• Future Prospects
– Wireless, implanted EMG sensors incorporate functional advantages of
wire electrodes with minimal invasiveness.

This article and any supplementary material should be cited as follows: Scheme E, Englehart K. Electromyogram pattern recognition
for control of powered upper-limb prostheses: State of the art and challenges for clinical use. J Rehabil Res Dev. 2011;48(6):643–60.
DOI:10.1682/JRRD.2010.09.0177

You might also like