EMG Decomposition Tutorial

*Motor units

*The EMG signal

*EMG Decomposition

A Brief Introduction to EMG Decomposition

Motor units.

Muscles are made up of slender fibers about 50 microns in diameter—about the same thickness as a human hair. These fibers are organized into groups known as motor units. All the fibers in a motor unit are innervated by a single motoneuron, and so they act together during a muscular contraction. The nervous system activates the motor unit by sending electrical impulses along the motoneuron axon. Each impulse causes the fibers to twitch, and when the impulses come at a fast enough rate---about fifteen per second---the twitches fuse to produce a steady force.

The nervous system controls the force of a muscular contraction by turning motor units on and off (recruitment) and by modulating their discharge rates (rate coding). A muscle's motor units are recruited in a fixed order from smallest to largest (the "size principle"). During a sustained contraction, discharge rates typically range from ten to twenty-five discharges per second. The intervals between the discharges of a particular motor unit are not perfectly regular, but can vary by several milliseconds from one discharge to the next. The discharge times of different motor units are largely independent of one another.

Each nerve impulse triggers an electrical discharge, or action potential, in each of the muscle fibers that it innervates. The action potentials originate at the point of innervation and then spread along the muscle fibers in two waves outward from the inntervation point toward the two ends of the fiber. These waves activate the contractile apparatus inside the muscle fibers. These waves also produce an electrical potential in the surrounding muscle tissue that can be detected by a needle or fine-wire electrode. This is the basis of electromyography.

When a motor unit discharges, the electrical potentials from all the muscle fibers of the motor unit sum together to produce a compound potential called the motor-unit action potential (MUAP). MUAPs typically last from one to several milliseconds, although their exact size and shape depend on the where the electrode is located with respect to the fibers. MUAPs from different motor units tend to have different shapes, while MUAPs recorded by the same electrode from the same motor unit have more or less the same shape from discharge to discharge.

The EMG signal.

The EMG signal is the summation of the discharges of all the motor units within the pick-up range of the electrode. During a sustained weak contraction, a needle electrode might detect two motor units discharging independently at rates of about ten discharges per second. The EMG signal will consist of two distinct trains of MUAPs. Most of the MUAPs will be clearly recognizable. Occasionally, though, the two motor units may discharge at nearly the same time, and the two MUAPs will overlap one another. This is called a superposition. The resultant waveform can be either larger or smaller than the individual MUAPs, depending on whether the overlap results in constructive or destructive interference.

During a somewhat stronger contraction, the EMG signal might contain MUAP trains from four to eight nearby motor units and several more distant ones, with discharge rates ranging up to fifteen discharges per second. The individual MUAPs will be difficult to sort out because of the high number of superpositions. The MUAPs from the nearby motor units can be enhanced by high-pass filtering. This accentuates their spikes, making them sharper and narrower, and, at the same time, attenuates the duller, broader MUAPs from the more distant motor units. Even though the spikes from the nearby motor units are narrow, they tend to retain their individuality. In fact, different motor units can usually be distinguished by their spikes more reliably than by their unfiltered MUAPs.

The EMG signal from an even stronger contraction might contain so much activity that even after high-pass filtering the individual MUAP trains cannot be identified with any degree of reliability. The complexity of an EMG signal depends on the strength of the contraction, whether the strength remains constant or not (isotonic or non-isotonic), whether the length of the muscle remains constant or not (isometric or non-isometric), and the type of electrode.

The three main categories of electrodes used to record EMG signals are needle electrodes, fine-wire electrodes, and surface electrodes. Needle electrodes have the advantage that they can be manipulated within the muscle to sample different parts of the muscle and to optimize signal characteristics. Fine-wire electrodes are typically inserted with a hypodermic needle. The ends of the wires remain in the muscle while the wires pass through the skin to connect to the amplifier. One advantage of fine-wire electrodes is that they tend to remain in place well throughout a long experiment.

The selectivity of needle and fine-wire electrodes depends on the size of the recording surface. The concentric and monopolar needle electrodes used in clinical EMG examinations record primarily from just the tip of the needle. These electrodes sample a region about half a millimeter in radius, which can include muscle fibers from dozens of motor units. The same is true for fine-wire electrodes with a half-millimeter-long recording surface. With these electrodes it is usually possible to sort out individual MUAP trains from signals recorded at up to about a third of maximum strength in many muscles. More selective electrodes with smaller recording surfaces have been developed for detecting individual MUAP trains from stronger contractions.

Surface electrodes have the advantage that they are completely non-invasive, but the limitation that they can only sample superficial muscles. Since surface electrodes are far away from the muscle fibers, the MUAPs they record are small in amplitude and tend to all look alike. Electrode arrays are often necessary in order to obtain decomposable signals.

EMG Decomposition

The process of sorting out the individual MUAP trains in an EMG signal is called EMG decomposition. Since each MUAP is related in a one-to-one way with the discharge of a motoneuron, EMG decomposition provides a unique way to observe the behavior of individual motoneurons in the intact human nervous system. Also, since the shapes of the MUAPs convey information about the characteristics and arrangement of the muscle fibers, EMG decomposition providing a unique way to study motor-unit organization in intact human muscles. This information is also used in clinical neurophysiology for diagnosing neuromuscular disorders.

Before the advent of computers, EMG signals were sometimes recorded photographically from the oscilloscope screen and then decomposed manually by marking the repetitive occurrences of each distinct spike. Nowadays computers are used, but the process is largely the same. It involves three main steps.

First, the shapes of the different MUAPs must be determined. This can be done by sorting the spikes in the signal on the basis of their shapes. Sorting will reveal clusters of spikes with similar shapes as well as some spikes with shapes that are different from all the others. Spikes with similar shapes are likely to be different discharges from the same motor units, while spikes with unique shapes are likely to be chance superpositions. In this way it is usually possible to work out the number of different MUAPs and to establish templates of their shapes.

The second step is to try to determine the source of every spike in the signal. Many spikes can be easily recognized as a discharge of one motor unit or another. The constituents of superpositions can be more difficult to work out. If the overlap is only slight, the constituents might still be recognizable. If the overlap is complete it might be necessary to try different alignments of the templates to see which gives the closest fit. The motor-unit discharge patterns can also be used to help determine which motor units are involved. Since these patterns tend to be fairly regular, the approximate timing of a particular discharge can be estimated from the timing of the preceding or following one.

The final step in decomposition is to check the results. If there are gaps or uneven intervals in any of the discharge patterns, or if there are spikes in the signal that have not been accounted for, then the decomposition is probably not correct. On the other hand, if all the activity in the signal has been adequately accounted for by a set of motor units with physiologically realistic discharge patterns, then there is a good chance that the decomposition is substantially complete and correct.

Of course some signals, and some MUAP trains within those signals, can be decomposed more reliably than others. Decomposability depends on several factors including the complexity of the signal, the level of background noise, the variability of the MUAPs from the same motor units, and the similarity of the MUAPs from different motor units. Some signals can be decomposed with a high degree of certainty, while others cannot be reliably decomposed at all.

Veterans Affairs Palo Alto Rehabilitation Research and Development Center
National Institute of Neurological Disorders and Stroke