Page 311 - Biomedical Engineering and Design Handbook Volume 1, Fundamentals

288 BIOMECHANICS OF THE HUMAN BODY

patterns to reproduce coordinated movements and estimate muscle forces (Neptune et al., 2001;
Thelen and Anderson, 2006). In these studies, the cost function was the error between measured and
simulated joint kinematics and ground reaction forces.
Unfortunately, the linear cost functions described above have been shown to have difficulty
predicting cocontraction and recorded muscle activation patterns, and cannot account for different
recruitment patterns during various tasks (Herzog and Leonard, 1991; Buchanan and Shreeve, 1996).
The dynamics-simulation-based cost functions have limitations because they do not account for the
differences in neural control strategies that may be employed by different people. They generally
ignore cocontractions, which is a commonly used strategy when learning a new task or when injured.
Electromyography (EMG) is the study of the electrical signal recorded by electrodes over or
inside muscles. An EMG signal includes real-time information about the electrical activity of a
specific muscle—which is related to muscle force—and has been studied for over 50 years. The rela-
tionship between EMG and muscle force has been studied during isometric contractions and dynamic
contractions. It has been reported to be linear (Lippold, 1952; Bouisset and Goubel, 1971; Moritani
and deVries, 1978) and slightly nonlinear (Heckathorne and Childress, 1981; Woods and Bigland-
Ritchie, 1983) during isometric contractions. For dynamic contractions, the change of muscle force
has been found to be also dependent on the change of muscle fiber length (Gordon et al., 1966) and
fiber velocity (Edman, 1978; Edman, 1979), so the relationship between the amplitude of EMG and
the force output should be modeled more comprehensively.
Based on the research of muscle force-EMG relationship, different EMG-driven biomechanical
models have been developed to estimate muscle forces (De Duca and Forrest, 1973; Hof and Van den
Berg, 1981; Buchanan et al., 1993; Thelen et al., 1994; Lloyd and Buchanan, 1996; Lloyd and Besier,
2003; Buchanan et al., 2004; Buchanan et al., 2005). Since these models determine muscle forces
based on recorded EMG data, they can account for different muscle activation patterns and may
predict muscle force more accurately (especially when studying different neural control strategies)
than other methods.

12.2 THE EMG SIGNAL

12.2.1 How EMG Is Generated?
The EMG signal is a complex, time-varying biopotential waveform, which is emanated by the muscle
underneath the electrode. It includes information about muscle activity and has been used as a
primary tool to study muscle function. Clinicians can diagnose problems in the neuromuscular system
by analyzing the onset/offset of EMG signal, or the peak amplitude of EMG signal; biomechanists
can use the EMG signal as an input to estimate muscle force; neurophysiologists can use the EMG
signal to identify different mechanisms of motor control and learning. In this section we will intro-
duce how muscle force is generated and EMG signal is detected (Enoka, 2002).
In human skeletal muscle, the muscle fibers do not contract individually; instead, they act as
small groups in concert. A single smallest controllable muscular unit is named a motor unit. A motor
unit is composed of a single motor neuron, the multiple branches of its axon, and the muscle fibers
that the motor neuron innervates (Fig. 12.1). Voluntary contraction of a motor unit initiates from an
action potential (an all-or-none electrical impulse that is issued by a cell if the input to the cell
exceeds its threshold) sent out from the central nervous system (CNS), and travels down the axon to
the muscle fibers. When the impulse of action potential reaches the muscle fibers, it activates all the
fibers in the motor unit almost simultaneously.
A motor neuron is an efferent neuron that transmits the output signal (action potential) from the
CNS to skeletal muscle. There are two main kinds of motor neurons: upper motor neurons and lower
motor neurons. An upper motor neuron resides in the motor region of cerebral cortex or the brain
stem, and extends its axon down the spinal cord to the lower motor neuron. The lower motor neuron
then sends its axon out through the ventral root of the spinal cord, and the axon is bundled together
into peripheral nerves that reach the target muscle. When the axon reaches the muscle, it splits into