Biomedical Engineering Reference
In-Depth Information
between this reset, integrated curve and the linear envelope. Both follow the
trend of muscle activity. If the reset time is too high, it will not be able to
follow rapid fluctuations of EMG activity, and if it is reset too frequently,
noise will be present in the trendline. If the integrated peaks are divided by
the integration time, the amplitude of the signal can again be reported in
millivolts.
A third common form of integration uses a voltage level reset. The integra-
tion begins before the contraction. If the muscle activity is high, the integrator
will rapidly charge up to the reset level; if the activity is low, it will take
longer to reach rest. Thus, the strength of the muscle contraction is measured
by the frequency of the resets. A high frequency of reset pulses reflects high
muscle activity; a low frequency represents low muscle activity. Intuitively,
such a relationship is attractive to neurophysiologists because it has a similar-
ity to the action potential rate present in the neural system. The total number
of counts over a given time is proportional to the EMG activity. Thus, if the
threshold voltage level and the gain of the integrater are known, the total
EMG activity (millivolt-seconds) can be determined.
10.4 RELATIONSHIP BETWEEN ELECTROMYOGRAM AND
BIOMECHANICAL VARIABLES
The major reason for processing the basic EMG is to derive a relationship
between it and some measure of muscle function. A question that has been
posed for years is: “How valuable is the EMG in predicting muscle tension?”
Such a relationship is very attractive because it would give an inexpensive
and noninvasive way of monitoring muscle tension. Also, there may be infor-
mation in the EMG concerning muscle metabolism, power, fatigue state, or
contractile elements recruited.
10.4.1 Electromyogram versus Isometric Tension
Bouisset (1973) has presented an excellent review of the state of knowledge
regarding the EMG and muscle tension in normal isometric contractions.
The EMG processed through a linear envelope detector has been widely
used to compare the EMG-tension relationship, especially if the tension is
changing with time. If constant tension experiments are done, it is sufficient
to calculate the average of the full-wave rectified signal, which is the same
as that derived from a long time-constant linear envelope circuit. Both linear
and nonlinear relationships between EMG amplitude and tension have been
discovered. Typical of the work reporting linear relationships is an early study
by Lippold (1952) on the calf muscles of humans. Zuniga and Simons (1969)
and Vredenbregt and Rau (1973), on the other hand, found quite nonlinear
relationships between tension and EMG in the elbow flexors over a wide
range of joint angles. Both these studies were, in effect, static calibrations of
Search WWH ::
Custom Search