GLENN S. KASMAN, MS, PT
Surface electromyography (SEMG) is the recording of muscle action potentials with skin surface electrodes. This article summarizes the rationale for incorporation of SEMG in evaluation and treatment programs for patients with musculoskeletal pain syndromes. Details on instrumentation, recording technique, limitations, and applications for specific clinical disorders can be found elsewhere (Cram and Kasman, 1998; Kasman et al., 1998).
Physiologic Rationale and Signal Processing
Motor activity is subserved by commands that are generated in the central nervous system and transmitted along alpha motor neurons to the periphery. Following chemical transmission across the neuromuscular junction, action potentials are produced along the sarcolemma, and electrical excitation becomes coupled to sarcomere shortening via complex chemical and micromechanical processes. Fundamentally, electrodes placed in the vicinity of excitable membranes will detect action potential events. SEMG electrodes detect the algebraic sum of voltages associated with muscle action potentials within their detection zone (Basmajian and De Luca, 1985). The SEMG signal represents the relative level of recruitment of an ensemble of motor units that underlay the electrodes.
Basics of the SEMG system have been described for clinicians in numerous sources (Basmajian and De Luca, 1985; Basmajian, 1989; Cram and Kasman, 1998; Peek, 1995; Soderberg, 1992; Turker, 1995). Electrodes are usually in the shape of 0.5-1.0 cm discs coated with silver-silver chloride. Some SEMG configurations require a paste or gel to be placed as a conductive medium between the electrode detection surfaces and the skin whereas others can be used “dry”. Each recording channel is composed of two active electrodes and a reference electrode. Active electrodes tend to be spaced with their centers 1.0-3.0 cm apart. The difference in electrical charge between each active electrode and the reference makes for inputs to a differential amplifier with high input impedance. One of the amplifier input signals is inverted. This process has the effect of canceling elements of the signal that are common to both inputs (typically unwanted noise and artifacts) and passing muscle voltage components for amplification.
The signal is next subjected to frequency bandpass filtering to enhance the signal to noise ratio. Surface voltages resulting from muscle action potentials can be decomposed into a specific frequency spectrum. Filters are used to pass frequencies related to muscle activity and to reject frequencies that are associated with noise.
The SEMG signal may sometimes be processed further to ease interpretation of a visual or auditory display. Processing often includes full wave rectification so that the plus-minus variations of the wave form are converted into a unidirectional signal. Several methods exist to smooth the peaks and valleys of the rectified wave form to ease inspection as well as to quantify the amplitude of the processed muscle signal.
Assessment of patients using SEMG
The amplitude of the SEMG signal is usually expressed as some number of microvolts, noted as series of relatively instantaneous measurements, or averaged or integrated over a clinically meaningful period of time. Amplitude analyses are conducted to evaluate the magnitude and timing pattern of muscle activity. Inferences are drawn regarding a muscle’s role in effecting a particular posture or movement, and how that role is altered by pathologic processes. The SEMG activity of a homologous muscle pair or that of an agonist, compared with it’s antagonists or synergists, is examined to assess muscle balance.
Imbalance occurs when the relative stiffness of muscles that participate in concert to execute a specific movement is inappropriately coordinated (Kasman et al., 1998). Muscle imbalance is presumably a function both of faulty central nervous system motor control and peripheral factors such as inefficient length-tension relationships and passive myofascial compliance. SEMG studies may therefore provide insight into the active component of muscle imbalance and can be linked by clinicians to the results of physical examination. Untoward motor programming may be influenced by nociception, perception, affect, beliefs, metabolic and nutritional issues, segmental and suprasegmental motor reflexes, sympathetically mediated reflexes, and a host of factors related to articular function and periarticular connective tissues. Analysis with SEMG can help clinicians in identifying relationships between muscle impairments and other physical and pyschologic impairments. Classification of impairments with observed functional limitations and disabilities can then be used to drive treatment planning in a thoughtful way (Jette, 1996). The effects of specific treatment procedures on muscle dysfunction also can be objectively verified, quantified, and documented with SEMG. Examination of SEMG amplitudes has been described for intervention with a wide variety of musculoskeletal disorders (Cram and Kasman, 1998; Kasman et al., 1998).
Clinically less common than amplitude analyses, investigation in the frequency domain is performed to study muscular fatigue. The frequency spectrum of the SEMG signal shifts in a reliable way with fatigue (Basmajian and De Luca, 1985). That is, the frequency spectrum becomes compressed toward slower values due to neuromuscular and metabolic changes associated with high intensity isometric contractions. The shift begins as the contractions are sustained beyond a short time, preceding the actual loss of force, and continues as force declines. This means of fatigue monitoring may have certain advantages over other measures (Ng, 1997) and successfully discriminates spinal pain patients from control subjects with impressive accuracy (Klein et al., 1991; Roy et al., 1995; Gogia and Sabbahi, 1990).
SEMG feedback training
In addition to clinical and kinesiological evaluations, the SEMG display is often used as a means of feedback for motor learning by patients (Kasman et al., 1998). Muscle cues produced by a SEMG device are far richer than those derived from a subject’s intrinsic sensory apparatus. Initially, a patient may have little idea how to change the activity of a muscle that is not under intuitive voluntary control. The patient may not possess a suitable motor programming scheme to achieve the goal (for example, increased activation of one muscle relative to another) and may have difficulty distinguishing correct performance from error. Cues on the SEMG display are obvious and serve as a reference of correctness. Thus the patient becomes able to evaluate various motor strategies for those that meet the goal. Successful strategies are repeated and ineffective strategies are discarded. The patient identifies a progressively smaller subset of effective motor behaviors over time. SEMG feedback is used cognitively to label subtle intrinsic sensations as indicative of changes in muscle activity. Through the repeated association of artificial, extrinsic cues from the SEMG machine with natural kinesthetic sensations, an intrinsic reference of correctness is formed. The learner forms mature sensory identification and motor programming schema, and can then achieve the goal independently.
The clinical objectives of feedback training with SEMG are relatively straightforward. Patients with muscle hyperactivity use feedback cues to reduce muscle output. For example, a patient with neck pain and upper trapezius hyperactivity could attend to the SEMG display to help improve posture, self-regulate responses to emotional stressors, or identify ergonomic improvements and motor skills for the workplace. A different patient with headaches and temporomandibular pain might produce chronic masseter and temporalis hyperactivity associated with chronic jaw clenching. Specific SEMG feedback techniques could be used to promote kinesthetic awareness, muscle relaxation, and reduction of parafunctional behaviors involving the temporomandibular region.
Patients with muscle hypoactivity incorporate SEMG feedback while learning to increase muscle recruitment. For example, a patient might show quadriceps inhibition after knee surgery that delays progress along a standardized clinical pathway. That patient could watch a SEMG display as his or her post-operative exercises are performed. Exercise variants, cognitive strategies, and adjunctive therapeutic agents would be trialed for those that facilitate quadriceps activity. Successful techniques would then be repeated while the patient attempts to raise the SEMG amplitude to match a goal marker on the display, set to progressively higher microvolt values over time.
In addition to training greater and lesser muscle responses as separate objectives, patients may learn to simultaneously increase the activity of a hypoactive muscle while decreasing that of a hyperactive muscle. This coordination training takes place between an agonist with it’s antagonists or synergists. For example, the patient alluded to previously with neck pain and upper trapezius hyperactivity might also show hypoactivity of the lower trapezius. This patient would try to raise the amplitude of the lower trapezius signal, and decrease the amplitude of the upper trapezius signal, during arm elevation maneuvers and simulated functional tasks. Successful training would presumably result in better muscle balance for upward scapular rotation and stabilization, leading to improved biomechanical relationships throughout the neck and shoulder girdle.
Advantages of SEMGas a means of muscle monitoring
SEMG techniques offer distinct conveniences compared with other means of muscle monitoring. The methods are noninvasive and painless. Hence use of SEMG tends to be readily accepted by patients and is generally quite safe. Although lead wires are used to connect the electrodes with the main instrument body (telemetry systems can be substituted if necessary), patients routinely are free to assume any position that is desired, including those for functional tasks. Recordings are feasible where dynamometers would be impractical, for example with investigation of facial muscles or selective examination of the vastus portions of the quadriceps. The SEMG display resolves changes in the magnitude and timing of muscle activity with far greater sensitivity than a clinician’s or patient’s eyes and hands. An entire range of activity levels can be captured for inspection, from voltages associated with activation of one or a few motor units to maximal effort recruitment. Within certain limits, the activity of particular muscles or muscle groups can be isolated. Set up becomes facile once the practitioner is experienced.
Like any clinical technique, SEMG has limitations. It is important to recognize that SEMG does not measure force, pain, anxiety, muscle length, joint position, or anything else other than voltage. With proper recording technique, the voltage pattern displayed with SEMG is representative of muscle recruitment. Inferences regarding clinical syndromes, however, are complicated by a complex interplay of neuromuscular, biomechanical, and psychological factors. Moreover, interpretation of SEMG activity can be subject to error brought about by certain effects of electrode configuration, tissue impedance, and other circumstances inherent to each recording set up. Thus clinicians who wish to perform SEMG procedures should become well versed with technical aspects of electrophysiological recording as well as models for clinical intervention.
The situation today
Technological advances have enabled commercial SEMG units to be miniaturized for ambulatory recordings of one to four channels of muscle activity. Patients can perform functional activities for a protracted time and the resultant SEMG data downloaded for analysis. Portable units are easily incorporated into therapeutic exercise programs in the clinic gym or prescribed for home programs. Commercial systems that incorporate a desk top computer are capable of simultaneous recordings from eight or more channels; sophisticated statistical processing of amplitude, timing and frequency variables; and a plethora of options for patient feedback. Software engineers continue to develop more powerful products while exploiting graphical user interfaces so that operation becomes simpler. Manufacturers and vendors are able to deliver SEMG products to consumers with a cost value that outstrips the pricing of earlier models.
Discussion in the community of health care providers who use SEMG extends to many patient populations. These include repetitive strain injuries in workers and athletes, dysfunction in patients who have sustained acute traumatic injuries, problems in patients with musculoskeletal dysfunction of insidious onset, and pain issues in patients with excessive psychophysiologic arousal. Numerous schools of thought can be found that embrace principles from psychology and movement science. Hence, SEMG can be regarded as a multidisciplinary modality. SEMG procedures should not be employed solely because a patient has chronic pain, but rather when aberrant muscle activity is suspected as being a primary contributory factor to dysfunction and evaluation with SEMG will impact treatment planning. Feedback training with SEMG may or may not then be appropriate to facilitate motor learning by the patient. The important point is that SEMG should be used to enhance functionally meaningful outcomes that reduce patient disability, in ways that support patient satisfaction, while controlling the financial and social costs of care.
With those objectives in mind SEMG may be considered, for example, with patients with:
- tension-type headache
- temporomandibular pain syndromes
- whiplash injuries
- neck pain associated with repetitive work tasks
- shoulder instabilities
- shoulder impingement syndromes
- peri-scapular pain syndromes
- lateral/medial epicondylalgia
- carpal tunnel syndrome
- post-surgical wrist and hand rehabilitation
- chronic lumbar dysfunction
- delayed rehabilitation after cervical/lumbar surgical fusion
- pelvic floor pain syndromes
- chronic hip dysfunction
- delayed rehabilitation after anterior cruciate ligament repair
- delayed rehabilitation after total knee replacement
- patellofemoral pain syndromes
- generalized muscle tension in psychophsyiologic stress syndromes
- hysterical muscle weakness or malingering
SEMG is additionally used to assist with rehabilitation efforts in patients recovering from severe neurologic insult, such as stroke, head injury, spinal cord injury, and peripheral nerve trauma. Urinary and fecal incontinence are also addressed with SEMG techniques as part of conservative intervention programs. Lastly, SEMG is combined with other forms of physiologic monitoring to teach self-regulation of autonomic dysfunction and as an adjunct in mental health counseling.
Since coming into clinical usage in the 1960’s, the number of scientific publications involving SEMG has grown nearly at an exponential rate (Cram, 1997). This trend in research is paralleled by developments in SEMG equipment and clinical procedures. Care providers may choose to make SEMG an integral part of their practice or reserve it’s use for occasional investigations of muscle activity and patient training. In any event, SEMG provides a unique means of monitoring muscle activity. Each clinician’s repertoire of skills may be broadened by inclusion of SEMG, while patients are provided with powerful opportunities for motor learning. It seems probable that the future will bring new applications of SEMG in performance enhancement in non-injured populations, new developments in forensic medicine, as well as refined approaches to SEMG with musculoskeletal and neuromuscular injuries.
- Basmajian JV, ed. Biofeedback: Principles and Practice for Clinicians, 3rd ed. Baltimore: Willimas and Wilkins, 1989.
- Basmajian JV, De Luca CJ. Muscles Alive: Their Functions Revealed by Electromyography, 5th ed. Baltimore, MD: Williams & Wilkins; 1985.
- Cram, JR. SEMG Field Growing. Biofeedback, newsmagazine of the Association for Applied Psychophysiology and Biofeedback (AAPB), Wheatridge, CO: AAPB. 1997; Fall: 11A.
- Cram JR, Kasman, GS. Introduction to Surface Electromyography. Gaithersburg, MD: Aspen Publishers, 1998.
- Gogia P, Sabbahi M. Median frequency of the myoelectric signal in cervical paraspinal muscles. Arch Phys Med Rehabil. 1990;71:408-414.
- Jette AM. Physical disablement concepts for physical therapy research and practice. Phys Ther. 1994;74:380-386.
- Kasman GS, Cram JR, Wolf SL. Clinical Applications in Surface Electromyography: Chronic Musculoskeletal Pain. Gaithersburg, MD: Aspen Publishers, 1998.
- Klein AB, Snyder-Mackler L, Roy SH, De Luca CJ: Comparison of spinal mobility and isometric trunk extensor forces with electromyographic spectral analysis in identifying low back pain. Phys Ther. 1991;71:445-454.
- Ng JKF, Richardson CA, Jull GA. Electromyographic amplitude and frequency changes in the iliocostalis and multifidus muscles during a trunk holding test. Phys Ther. 1977;77:954-961.
- Peek CJ. A primer of biofeedback instrumentation. In Schwartz M, ed. Biofeedback: A practitioner’s Guide. New York: Guilford Press; 1987:45-95.
- Roy SH, De Luca CJ, Emley M, Buijs RJC. Spectral electromyographic assessment of back muscles in patients with low back pain undergoing rehabilitation. Spine. 1995;20:38-48.l
- Soderberg GL, ed. Selected Topics in Surface Electromyography for Use in the Occupational Setting: Expert Perspective. Washington, DC: US Dept of Health and Human Services; 1992. US Dept of Health and Human Services publication NIOSH 91-100.
- Turker KS. Electromyography: Some methodological problems and issues. Phys Ther. 1993;73:698-710.