Insertion of a needle into a normal muscle injures and mechanically stimulates many muscle fibers, producing a burst of action potentials of short duration (< 300 msec) . At rest normal muscle is electrically silent as normal muscle tone is not the result of electrical contraction of muscle fibers. As an electrical impulse travels along the surface of a single muscle fiber toward the recording electrode, the impulse becomes positive (downward deflection shown on the computer screen by convention) relative to the reference electrode. As the impulse comes beneath the electrode, the waveform becomes negative (upward deflection) and then becomes slightly positive and returns to baseline as the impulse travels past the electrode (Fig. 3.2). A single muscle fiber contraction lasts about 2–4 msec and is less than 300 volt in amplitude. The firing of a single muscle fiber (called fibrillation which does not cause visible muscle movement) does not occur normally and is a sign of muscle membrane instability either from denervation or myopathy. In normal muscle, an electrical impulse travels from a spinal cord anterior horn neuron (lower motor neuron) along its axon to eventually innervate 10–1000 muscle fibers (called a motor unit). The number of muscle fibers innervated depends on the muscle with proximal limb muscles having the highest number of innervated muscle fibers. During mild voluntary muscle contraction, an entire motor unit fires almost simultaneously producing a motor unit action potential (MUAP). A typical MUAP has 3–4 excursions across the baseline (phases) and a maximum amplitude of 0.5–5 mV (Fig. 3.2) . The shape and duration of a given MUAP remain quite constant on repeat firings and generally appear different from other nearby MUAPs.

Immediately after complete nerve transection, the muscle is paralyzed, unexcitable by nerve stimulation, and electrically silent by EMG except for insertion potentials. Beginning 2–3 weeks after a muscle loses its innervation, spontaneous individual muscle fiber contractions may appear. The EMG demonstrates fibrillations and positive sharp waves (brief monophasic positive spikes). Until the motor unit completely degenerates, spontaneous firing of the MUAP also occurs (called a fasciculation which produces a visible muscle twitch). If the nerve damage is incomplete and occurred several months earlier, the denervated muscle fiber induces adjacent motor nerves to branch or sprout and send a nerve branch to re-innervate the denervated muscle fiber called sprouting. MUAPs suggestive of sprouting are of longer duration, contain more phases, and may be of higher maximum amplitude than normal.

Death or dysfunction of scattered muscle fibers results in MUAPs during voluntary muscle contraction that are of shorter duration and lower amplitude than normal (Fig. 3.3). Some MUAPs may be polyphasic from loss of synchronous firing. In myositis, there may be accompanying fibrillations due to inflammatory damage to adjacent motor nerve endings .

Motor nerve conduction velocity studies measure the velocity of the fastest motor nerve axons at various points along a peripheral nerve. Peripheral nerves can be stimulated to fire by application of an electrical impulse to the skin overlying the nerve. When a muscle contracts, its electrical signal can be detected by placing an electrode on the skin above the muscle belly. The muscle electrical signal is recorded, and the time from electrical stimulus to muscle contraction (latency) can be determined and displayed on an oscilloscope. A motor nerve velocity is determined as follows (Fig. 3.4). By moving the stimulating electrode along the nerve pathway, differing latencies in msec to muscle contraction are determined. By measuring the distance along the nerve pathway between two stimuli, one can divide the nerve distance in mm by the latency difference in msec to obtain the nerve velocity in meters per second. Normal motor velocity of the median and ulnar nerves is 50–60 m/sec and of the sciatic nerve is 40–50 m/sec. Slowing of the motor nerve velocity may reflect loss of myelin along the nerve (often causing slowing of velocities to 20–30 m/sec) or loss of the fastest motor nerves (lesser degree of velocity slowing). Slowing of a motor nerve may occur along the entire nerve pathway or at a localized point of nerve compression, such as the ulnar nerve at the elbow.

Evaluating sensory nerve function is more difficult as the normal signals are weaker and more diffuse following an electrical stimulus because the conduction velocities of different sensory axons vary considerably. Sensory nerves may be unmyelinated and conduct at ½–2 m/sec or thinly myelinated and conduct at 10–20 m/sec. The most common sensory nerve test determines the latency time from electrical stimulation of the skin of a finger to a skin recording electrode site over the median nerve just proximal to the wrist. A delayed median nerve sensory latency suggests compression of the nerve at carpal tunnel. The test is safe, mildly expensive, somewhat uncomfortable, and takes about 30 min.

Information about the function of the neuromuscular junction can be obtained from repetitive nerve stimulation studies. Placement of a skin recording electrode over the belly of a muscle and stimulating the motor nerve produce a compound muscle action potential (CMAP) . If the nerve stimulation is repeated, the CMAPs appear identical on the oscilloscope. In diseases of the neuromuscular junction, the amplitude of the CMAPs may decrease or increase. In myasthenia gravis and botulism , repetitive nerve stimulation produces a decremental response in the CMAP. The test is safe, inexpensive, somewhat uncomfortable, and takes about 15 min.

Occasionally, there are indications to evaluate the integrity of central conduction along major sensory pathways (visual, auditory, and peripheral sensory system) called evoked potentials. As noted above, actual conduction velocities cannot be obtained but central modality-specific latencies can. Evoked potential tests record computer averages of the EEG that is time-locked to repeated (100–500 trials) specific sensory stimuli such as sound, light, or electrical stimulation of peripheral nerve. The computer averaging reduces background EEG electrical activity to zero while enhancing the time-locked stimulus signal. Abnormalities are characterized by a delay for the time-locked signal average to reach its destination or distortions (usually a prolongation of the waveform and loss of signal amplitude). Sensory evoked potentials are safe, inexpensive, and comfortable. The major indication is the evaluation of possible diseases that cause CNS demyelination of these sensory pathways.