EMG data is acquired from a contracting muscle by the use of one or more electrodes. Currently, the state of the art techniques use a concentric needle electrode (which has a diameter of about half a millimetre), which is placed into a muscle. The muscle is then isometrically contracted (that is, contracted without a change in length), allowing the electrical activity of the muscle to be recorded without causing a lot of pain to the subject.
By decomposing EMG data, it is possible to infer particular structural sources in the muscle responsible for individual observed potentials. These are called motor units, and the potentials associated with a particular motor unit are motor unit potentials, or MUPs.
Observation of a train of MUPs, makes it possible to relate quantitative statistical measures of the MUP shapes to the various motor units. The study of the techniques to do so are called quantitative electromyography, or QEMG.
My work focuses on the application of QEMG data to diagnostic decision making. Within QEMG data there is a great deal of information which is applicable to the diagnostic process and which can be used to infer the type and degree of involvement of disease. Equally important goals in such decision making are the well-established goal of ascertaining the most-likely association, as well as the establishing the degree of confidence which may be associated with a particular association.
An examination of decision confidence is particularly of interest in a clinical setting where decisions are made on collections of diagnostic samples; this leads to the question of whether additional samples will increase the confidence and accuracy of a diagnostic decision, and answers the basic question: “Should more data be collected before diagnosis?”
Recent results include the construction of a basic decision support system for the diagnosis of muscular disease, and current work involves the application and extension of this work into various types of muscular malady, and extension of the decision confidence concept and its visual representation.
Topics of immediate interest are the quantification of the presence and type of involvement for repetitive strain injuries, the generation of well-defined markers for myopathic and neuropathic disease state, further development of representations of decision confidence and visual metaphors for confidence-based clinical decision making.
EMGLAB
- Needle EMG muscle model program (Mac and Windows integration with Matlab
- Real and simulated signal data
- Decomposition algorithms
DQEMG
DQEMG is a collaboration with Dan Stashuk at the University of Waterloo. Integration of the DQEMG program with the Cadwell Laboratories Sierra Wave has been completed as the project DQEMGbridge.
Software and Data Sharing
We would like to share this software with other people interested in quantitative electromyography (QEMG). In particular, we would like to share data collected using this protocol with anyone interested in using DQEMG for analysis.
An Introduction and Overview document for DQEMG is available from this website.
f you are interested in obtaining and using DQEMG and the Sierra <=> DQEMG bridge, please email us at this address: dqemg@QEMG.org