The scheme for decomposing a myoelectric signal into its constituent motor unit action potential trains described in the paper [3] requires interaction from the human operator. In this paper, guidelines to be employed by the operator in assisting the computerized algorithms in identifying (classifying) a motor unit action potential are presented. The accuracy of the decomposition scheme was evaluated by decomposing a mathematically synthesized myoelectric signal. This signal was constructed by linearly superimposing eight mathematically generated motor unit action potential trains along with Gaussian noise. A skilled operator was able to decompose this signal with an accuracy of 99.8 percent, incurring one error in a total of 435 classifications. The decomposition reproducibility was evaluated by having two experienced operators independently decompose the same record of empirically obtained myoelectric signal. Their results were in total agreement for 479 motor unit action potential classifications belonging to five motor unit action potential trains. Up to eight motor unit action potential trains have been decomposed from one myoelectric signal.
[1]
E. Stålberg.
Single fibre electromyography
,
1979,
Trends in Neurosciences.
[2]
Carlo J. De Luca,et al.
Physiology and Mathematics of Myoelectric Signals
,
1979
.
[3]
Ronald S. Lefever,et al.
A Procedure for Decomposing the Myoelectric Signal Into Its Constituent Action Potentials - Part I: Technique, Theory, and Implementation
,
1982,
IEEE Transactions on Biomedical Engineering.
[4]
E Stålberg,et al.
The electromyographic jitter in normal human muscles.
,
1971,
Electroencephalography and clinical neurophysiology.
[5]
E. Stålberg,et al.
Single Fibre Electromyography for the Study of the Microphysiology of the Human Muscle
,
1973
.
[6]
Ronald Stanton LeFever.
Statistical analysis of concurrently active human motor units
,
1980
.