YALE JoUrNAL oF BIoLoGY AND MEDICINE 85 (2012), pp.201-215.
`Copyright © 2012.
`
`FoCUS: BIoMEDICAL ENGINEErING
`
`neuromuscular Electrical Stimulation for
`Skeletal Muscle Function
`
`Barbara M. Douceta, Amy Lamb, and Lisa Griffinb*
`
`aUniversity of Texas Medical Branch, Division of Rehabilitation Sciences, Galveston,
`Texas; bUniversity of Texas, Department of Kinesiology and Health Education, Austin,
`Texas
`
`Lack of neural innervation due to neurological damage renders muscle unable to produce
`force. Use of electrical stimulation is a medium in which investigators have tried to find a way
`to restore movement and the ability to perform activities of daily living. Different methods of
`applying electrical current to modify neuromuscular activity are electrical stimulation (ES†),
`neuromuscular electrical stimulation (NMES), transcutaneous electrical nerve stimulation
`(TENS), and functional electrical stimulation (FES). This review covers the aspects of elec-
`trical stimulation used for rehabilitation and functional purposes. Discussed are the various
`parameters of electrical stimulation, including frequency, pulse width/duration, duty cycle, in-
`tensity/amplitude, ramp time, pulse pattern, program duration, program frequency, and mus-
`cle group activated, and how they affect fatigue in the stimulated muscle.
`
`introduction
`
`Damage to the human nervous system
`during an event such as stroke or spinal
`cord injury (SCI) produces a rapid dener-
`vation of muscle resulting in weakness or
`paralysis. This lack of neural innervation
`renders muscle unable to produce the vol-
`untary forces needed to create joint move-
`
`ment that will allow functional perform-
`ance of daily tasks [1]. Numerous scientific
`investigations have focused on devices,
`strategies, and regimens that may poten-
`tially restore body movement critically
`needed for daily function and quality of
`life.
`
`Using electrical stimulation to produce
`human movement is not a novel procedure.
`
`*To whom all correspondence should be addressed: Lisa Griffin, PhD, Department of Ki-
`nesiology and Health Education, 222 Bellmont, 1 University Station, D3700, University of
`Texas at Austin, Austin, TX, 78712; Tele: 512-471-2786; Fax: 512-471-8914; Email:
`l.griffin@mail.utexas.edu.
`
`†Abbreviations: CFT, constant frequency trains; DFT, doublet frequency trains; ES, elec-
`trical stimulation; FES, functional electrical stimulation; NMES, neuromuscular electrical
`stimulation; SCI, spinal cord injury; TENS, transcutaneous electrical nerve stimulation;
`VFT, variable frequency trains.
`
`Keywords: functional, paralysis, rehabilitation, spinal cord injury, stroke
`201
`
`LUMENIS EX1016
`Page 1
`
`

`

`202 Doucet et al.: Neuromuscular Electrical Stimulation for Skeletal Muscle Function
`
`In 1790, Luigi Galvani first observed mo-
`tion after applying electrical wires to leg
`muscles severed from the body of frogs, and
`in 1831, Michael Faraday showed that elec-
`trical currents could stimulate nerves to cre-
`ate active movement [2]. One of the earliest
`clinical experiments that used electrical
`stimulation for muscle function stimulated
`the peroneal nerve in the leg in an effort to
`correct foot drop in persons with stroke-re-
`lated hemiplegia during ambulation [3].
`Whether used alone to improve motor
`impairment or embedded within complex
`systems to create functional multi-joint
`movement, the potential that electrical stim-
`ulation holds for rehabilitation recovery is
`immeasurable. Electrical stimulation is cur-
`rently used in many forms to facilitate
`changes in muscle action and performance.
`In clinical settings, electrical stimulation can
`be used for improving muscle strength, in-
`creasing range of motion, reducing edema,
`decreasing atrophy, healing tissue, and de-
`creasing pain. Neuromuscular electrical
`stimulation (NMES), used interchangeably
`with electrical stimulation (ES), is typically
`provided at higher frequencies (20-50 Hz)
`expressly to produce muscle tetany and con-
`traction that can be used for “functional”
`purposes and can be found in literature as
`early as 1964 [4]. TENS is an alternate form
`of electrical stimulation that historically
`used high frequencies for pain relief [5] but
`is now also administered at very low fre-
`quencies (sensory level TENS, 2-10 Hz) [6].
`TENS propagates along smaller afferent
`sensory fibers specifically to override pain
`impulses. When very low frequencies are
`used, TENS specifically targets sensory
`nerve fibers and does not activate motor
`fibers; therefore, no discernible muscle con-
`traction is produced.
`The acronym FES (functional electrical
`stimulation) is probably the most commonly
`used in the literature; however, a distinction
`should be made that this method of electrical
`stimulation usually refers to the process of
`pairing the stimulation simultaneously or in-
`termittently with a functional task as initially
`described by Moe and Post [7]. For exam-
`ple, Thrasher et al. [8] designed a program
`
`of FES for the upper extremity of persons
`with stroke that consisted of initial stimula-
`tion of the anterior and posterior deltoid, fol-
`lowed by triceps brachii stimulation. This
`resulted in flexion of the shoulder and elbow
`extension to produce a forward reaching
`motion for function. The second phase of the
`study stimulated wrist extensors and finger
`flexors to contract the fingers around an ob-
`ject in order to facilitate a grasping task. The
`stroke group that received FES in addition
`to conventional therapy significantly im-
`proved in function when compared to those
`receiving only conventional therapy. FES
`has also been used extensively to reproduce
`the activation pattern of lower extremity
`muscles to produce human gait [9] and to
`create the sequence of lower extremity mus-
`cle activation needed during a cycling task
`[10-12] in persons unable to actively per-
`form these movements. Several studies
`demonstrate the benefit of pairing ES with
`tasks that demand the use of intact cognitive
`and motor skills of the patient as compared
`to using ES simply as a passively delivered
`modality [13-16]. The term sometimes used
`to describe stimulation that cycles on and off
`repetitively without patient involvement is
`known as “cyclic” electrical stimulation
`[17,18].
`A significant limitation of any non-
`physiologically induced muscle activation is
`the overall decreased efficiency of contrac-
`tion and propensity for development of neu-
`romuscular fatigue. With NMES,
`the
`primary causes are suggested to be an alter-
`ation of the normal recruitment order and the
`unnatural simultaneous activation of motor
`units (see following section “Limitations of
`Electrical Stimulation”). Therefore, strate-
`gies must be designed as part of electrical
`stimulation regimens to offset the high de-
`gree of fatigue associated with ES.
`The delivery of electrical stimulation
`can be customized to reduce fatigue and op-
`timize force output by adjusting the associ-
`ated
`stimulation parameters. A
`full
`understanding of the settings that govern the
`stimulation is vital for the safety of the pa-
`tient and the success of the intervention.
`Consideration should be given to the fre-
`
`LUMENIS EX1016
`Page 2
`
`

`

`Doucet et al.: Neuromuscular Electrical Stimulation for Skeletal Muscle Function
`
`203
`
`quency, pulse width/duration, duty cycle, in-
`tensity/amplitude, ramp time, pulse pattern,
`program duration, program frequency, and
`muscle group activated.
`
`ParaMEtErS oF ElEctrical
`StiMulation
`
`Frequency
`
`Frequency refers to the pulses produced
`per second during stimulation and is stated
`in units of Hertz (Hz, e.g., 40 Hz = 40 pulses
`per second). The frequencies of electrical
`stimulation used can vary widely depending
`on the goals of the task or intervention, but
`most clinical regimens use 20-50Hz patterns
`for optimal results [19,20]. In order to avoid
`fatigue or discomfort, constant low fre-
`quency stimulation is typically used, which
`produces a smooth contraction at low force
`levels [21]. In a study comparing several dif-
`ferent frequencies and stimulation patterns,
`frequencies under 16Hz were not sufficient
`to elicit a strong enough contraction to allow
`the quadriceps to extend to a target of 40º
`[22]. Interestingly, lower frequencies of
`stimulation have been shown to impart a
`long-lasting depression of force output
`known as “low-frequency fatigue,” first de-
`scribed by Edwards, Hill, Jones, and Merton
`(1977). These researchers observed that fa-
`tigued muscle stimulated with lower fre-
`quencies (10-30Hz) had the potential to
`produce lower forces, a condition that lasted
`for 24 hours or longer; the same effect was
`not seen when the muscle was stimulated
`with higher frequencies. Later work by
`Bigland-Ritchie, Jones, and Woods (1979)
`showed that higher frequencies of stimula-
`tion (50 Hz and 80 Hz) administered to hand
`muscles resulted in a rapid decline in force
`after approximately 20s. More recently,
`stimulation frequency rates closely aligned
`with physiological rates of motor unit dis-
`charge were studied in the hand that showed
`a consistent frequency of 30 Hz preserved
`force better than a decreasing frequency pat-
`tern (30 Hz decreasing to 15 Hz) [23]. Mang
`et al. [24] showed that high frequencies of
`peripheral stimulation can have central con-
`
`tributions as well; activation of motor neu-
`rons in the spinal pool was highest when the
`tibialis anterior muscle was stimulated with
`100Hz as compared to stimulation at 10 and
`50 Hz. Higher frequencies are generally re-
`ported to be more comfortable because the
`force response is smoothed and has a tin-
`gling effect, whereas lower frequencies elicit
`a tapping effect where individual pulses can
`be distinguished [6].
`
`raMPing oF StiMulation
`FrEquEncy
`
`Frequently, a gradation of stimulation
`up to the desired frequency and intensity is
`used for patient comfort. Ramp time refers
`to the period of time from when the stimu-
`lation is turned on until the actual onset of
`the desired frequency [25]. Ramp time is
`used in clinical applications when a patient
`may have increased tone that creates resist-
`ance against the stimulated movement. For
`instance, a person with flexor hypertonicity
`at the elbow would benefit from a gradual
`ramping up of stimulation frequency to
`allow more time to activate elbow extensors
`moving in opposition to tightened flexors to
`successfully complete the movement [26].
`Ramp times of 1 to 3 seconds are common
`in rehabilitation regimens with longer ramp
`times sometimes used for hypertonic or
`spastic musculature or for the patient with
`an increased sensitivity to stimulation [25].
`Ramp times also can be modulated in multi-
`ple-muscle applications such as standing
`and walking to produce smooth gradations
`of tetany between individual muscles and
`more closely replicate natural movement
`[27].
`
`PulSE Width/duration
`
`Electrical stimulation devices deliver
`pulses in waveform patterns that are often
`represented by geometric shapes such as
`square, peaked, or sine wave. These shapes
`characterize electrical current that rises
`above a zero baseline for the extent of the
`stimulation paradigm (uniphasic; e.g., direct
`current) or current that alternates above and
`
`LUMENIS EX1016
`Page 3
`
`

`

`204 Doucet et al.: Neuromuscular Electrical Stimulation for Skeletal Muscle Function
`
`below the baseline (biphasic or alternating
`current) [28]. Biphasic and uniphasic wave-
`forms were noted to produce greater torque
`than polyphasic waveforms when adminis-
`tered to the quadriceps muscles of young
`healthy individuals [29].
`The time span of a single pulse is
`known as the pulse width or pulse duration.
`In biphasic (a positive phase combined with
`a negative) pulses, the pulse duration con-
`siders both phases [30]. Typically, dynamic
`quadriceps extensions similar to those used
`in FES cycling tests exhibit pulse widths be-
`tween 300µs-600µs [31-34]. Some investi-
`gators have suggested that low frequency
`stimulation with short pulse durations
`(500µs-1000µs) will exhibit a lower fatigue
`index [35]. However, even shorter pulse
`widths (10µs-50µs) have been shown to af-
`fect the recruitment of muscle fibers and can
`generate a larger maximum torque in a
`smaller number of fibers before causing a
`contraction in another muscle fascicle [36].
`This is important as a greater recruitment
`ratio within muscle fascicles can possibly in-
`crease performance time; therefore, pulse
`width can be increased to potentially recruit
`more fibers in the surrounding area as fa-
`tigue ensues. Recent work comparing 50,
`200, 500, and 1000µs pulse widths when 20
`Hz stimulation was delivered to the soleus
`muscle found that the wider pulse widths
`produced stronger contractions of plan-
`tarflexion and additionally augmented over-
`all contractile properties [37]. In addition,
`longer pulse durations will typically pene-
`trate more deeply into subcutaneous tissues,
`so these widths should be used when trying
`to impact secondary tissue layers [26].
`
`Duty Cycle
`
`Early work in persons with SCI demon-
`strated that when periods of force develop-
`ment were interrupted with silent periods,
`muscle tissue was able to recover more
`quickly and produce greater torque as com-
`pared to when constant stimulation patterns
`were used [38]. Cycling pulses on and off
`(intermittent stimulation) is a common prac-
`tice to preserve force development and si-
`multaneously increase comfort for the
`
`patient. Duty cycle describes the actual on
`and off time of an NMES program and is
`usually stated in ratio form, such as 1:2 (10
`seconds on, 20 seconds off) or percentages
`such as 70 percent, indicating time on per-
`centage when compared to total on and off
`time combined [25]. Common clinical ap-
`plications use a 1:3 duty cycle as standard,
`but this ratio can be modified to accommo-
`date the needs of the patient as well as the
`goals of the treatment [26].
`
`Amplitude/Intensity
`
`Another parameter that will contribute
`to fatigue is the strength of the current being
`administered or the intensity/amplitude
`(usually reported in milliamperes, mA) with
`which the stimulation is delivered. The
`higher the intensity, the stronger the depo-
`larizing effect in the structures underlying
`the electrodes [39]. Higher intensities can
`foster increases in strength; strength gains
`are consistently found following training
`with electrical stimulation programs [15,40-
`42]. Recent work examining the optimal pa-
`rameters for stimulation has suggested that
`lower intensities can induce more central
`nervous system input than higher intensities.
`Higher amplitudes of NMES activate a large
`number of muscle fibers that create forceful
`peripheral-mediated contractions, but an-
`tidromic transmission can occur (neural
`transmission toward the cell body rather
`than normal orthodromic transmission away
`from the cell body). Antidromic transmis-
`sion blocks both motor and sensory im-
`pulses emanating from the spinal motor
`pool, resulting in less overall CNS activa-
`tion [43]. The impact of stimulation ampli-
`tude on fatigue remains unclear. Downey et
`al. [44] found that when both frequency and
`amplitude were varied during a stimulation
`regimen of knee extension in healthy adults,
`more contractions were performed as com-
`pared to when a constant frequency and am-
`plitude program was used. In contrast, when
`NMES was delivered to the knee extensors
`of seven healthy participants and the influ-
`ence of frequency, pulse width, and ampli-
`tude on fatigue was studied, investigators
`found that fatigue decreased only when fre-
`
`LUMENIS EX1016
`Page 4
`
`

`

`Doucet et al.: Neuromuscular Electrical Stimulation for Skeletal Muscle Function
`
`205
`
`quency was decreased; lowering the other
`parameters had no appreciable effect on re-
`ducing fatigue [45]. Stimulation frequency
`rates closely aligned with physiological rates
`of motor unit discharge were studied in the
`hand that showed a consistent frequency of
`30 Hz preserved force better than a decreas-
`ing frequency pattern (30 Hz decreasing to
`15 Hz) [23]. Intensity will also factor into
`patient comfort with higher intensities being
`typically less tolerated; however, frequency
`and intensity inevitably will determine the
`quality of muscle contraction produced [25].
`
`StiMulation PulSE PattErnS
`
`Several investigations have examined
`the effects of various stimulation patterns on
`force output and neuromuscular fatigue.
`Common stimulation patterns studied are
`constant frequency trains (CFTs), variable
`frequency trains (VFTs), and doublet fre-
`quency trains (DFTs) [32-34,46-49]. CFTs
`are stimulation trains in which the frequency
`remains constant throughout the entire train.
`In contrast, VFTs are usually trains that
`begin with an initial doublet, (two closely
`spaced pulses, typically 5-10 µs apart) fol-
`lowed by pulses at a chosen frequency. The
`idea of VFT comes from studies where it
`was found that muscles have a “catchlike
`property,” a unique mechanical response to
`stimulation that allows muscle to hold a
`higher force level than normal (van Lun-
`teren, JAP 2000). This response enhances
`muscle tension prior to contraction when a
`brief, high frequency burst is followed by a
`train of subtetanic pulses [47,50,51]. The
`phenomenon does not appear to be a result
`of greater muscle fiber recruitment but an in-
`herent property of the individual muscle
`cells [50,52].
`In an isometric contraction of the thenar
`muscles of the hand, Bigland-Ritchie and
`colleagues showed that pulse trains that
`began with a doublet resulted in slower rates
`of force attenuation, suggesting a slower
`time to fatigue [53]. A similar study of iso-
`metric contraction of the thenar muscles of
`the hand examined variable patterns where a
`20Hz CFT fatigue task was compared to two
`
`other fatigue tasks; a 20Hz CFT was admin-
`istered for the first half of the fatigue task
`and then the frequency was increased grad-
`ually to 40Hz frequency or a 20Hz doublet
`train was added [54]. The findings of this
`study concluded that during submaximal
`stimulation, the doublet train was most ef-
`fective in producing higher average forces
`and force-time integrals. These studies pro-
`pose that using VFTs may be more benefi-
`cial in reducing fatigue in intrinsic hand
`muscles than CFTs alone.
`Other studies have observed the lower
`limb comparing CFTs, DFTs, and VFTs. In
`particular, one study fatigued the quadriceps
`muscle using CFTs and VFTs with varying
`interpulse intervals [52]. The fatigued mus-
`cle was then stimulated with either a CFT of
`14 or 18 Hz or a VFT (consisting of a train
`that used an initial doublet followed by a
`CFT). The results showed that VFT trains
`are more effective in producing higher peak
`forces, maintaining force output, and elicit-
`ing a more rapid rate of rise after being fa-
`tigued with a CFT as compared to using a
`VFT. Another investigation studied the ef-
`fect of using CFTs, VFTs, and DFTs with the
`same interpulse interval (50 ms, 20 Hz fre-
`quency) to elicit dynamic leg extension.
`DFTs had the best overall performance in
`time to reach target [55]. These findings sug-
`gest that there may be several optimal stim-
`ulation patterns, but these will be dependent
`on the task, population studied, and the mus-
`cle group being investigated.
`
`Electrode Placement
`
`The success of the FES current to reach
`underlying tissue is highly related to elec-
`trode size and placement, as well as the con-
`ductivity of the skin-electrode interface [56].
`In the past, a conductive gel was applied to
`the surface of electrodes to improve trans-
`mission of the current; typical stimulating
`electrodes used now are pre-gelled for con-
`venience. Larger surface electrodes will ac-
`tivate more muscle tissue but will disperse
`the current over a wider surface area, de-
`creasing current density. Smaller electrodes
`will concentrate current densities, allowing
`for focal concentration of current with less
`
`LUMENIS EX1016
`Page 5
`
`

`

`206 Doucet et al.: Neuromuscular Electrical Stimulation for Skeletal Muscle Function
`
`chance of stimulation crossover into nearby
`muscles, but dense current increases the
`chance for discomfort or pain [57]. Place-
`ment of electrodes will also markedly influ-
`ence the muscle response and should be
`carefully considered. Contention regarding
`optimal placement of electrodes is prevalent
`throughout the literature, with much of the
`debate centering on whether the muscle
`belly or the motor point is the preferential
`location. Rehabilitation therapists frequently
`place electrodes directly over the muscle
`belly [58] or in ineffective locations [59].
`Manufacturers also provide suggested elec-
`trode placement charts or guides that are
`usually included with the device purchase,
`also a source for clinicians using NMES in
`practice. A recent investigation of NMES
`delivered to the tibialis anterior and the vas-
`tus lateralis of the lower extremity compared
`electrode placement using the motor point
`of the muscle (accurately located through
`stimulation) with placement using the rec-
`ommended sites of several manufacturer's
`suggestions. This resulted in significant dif-
`ferences in muscle performance outcome;
`motor point placement not only produced
`higher torques, but blood flow and oxygen
`use was greater using the motor point posi-
`tions [60].
`
`quality of the skin-electrode interface and
`consistent placement of electrodes for re-
`peatability [61].
`
`doSing oF StiMulation
`
`Dosing of FES programs can vary
`greatly and will ultimately depend on the
`muscle being stimulated, parameters used,
`and overall goal of the intervention. A re-
`view of the use of FES for motor recovery of
`the upper extremity in stroke examined sev-
`eral investigations and found an array of
`dosing protocols used [20]. Program dura-
`tion ranged from 30 minutes one time per
`day to an hour at each session for three times
`per day. Overall period of treatment varied
`from 2 weeks to 3 months, with no justifi-
`cation by any author of why a particular dos-
`ing protocol was chosen. The researchers
`also found that increasing duration of treat-
`ment was not directly related to more suc-
`cessful outcomes; positive benefits were
`seen with short programs (2.5 hours/week),
`and limited benefits were seen with longer
`programs (21 hours/week). For rehabilita-
`tion of ambulation skills, FES-assisted walk-
`ing programs usually consist of three to five
`hour-long sessions per week for at least 4
`weeks [8].
`
`StiMulation intEnSity
`
`Stimulation can be delivered by means
`of constant voltage or constant current. The
`small portable units used in clinics and given
`to patients for home use are normally bat-
`tery-operated and have modifiable current
`settings usually delivered through a constant
`voltage system of approximately 150V.
`These units use transcutaneous surface elec-
`trodes that adhere to the skin and can be eas-
`ily removed. The contact area of the
`electrode is usually lined with the conduc-
`tive gel described earlier that facilitates
`movement of the current from the electrode
`into the skin. Because the units use alternat-
`ing current (AC) with a high degree of ad-
`justability, muscle activation through these
`devices can be sometimes be variable and
`inconsistent; outcomes will depend on the
`
`liMitationS oF ElEctrical
`StiMulation
`
`Although electrical stimulation has the
`capacity to produce movement in dener-
`vated, paralyzed, or spastic muscles, it is in-
`herently
`less
`efficient
`than human
`movement. Most importantly, NMES in-
`duces excessive neuromuscular fatigue. Re-
`searchers have studied frequency [31,34,62],
`pulse width [35,36,63], modulation of pulses
`[64], amplitude [63], electrode placement
`[65], and the use of variable frequency pulse
`patterns [22,52-55,66,67] to determine if fa-
`tigue can be reduced through a modification
`of any of these parameters.
`Causes for the excessive fatigue ob-
`served during NMES are multiple: First,
`NMES has the propensity to alter normal-
`motor unit recruitment order [68]. In normal
`
`LUMENIS EX1016
`Page 6
`
`

`

`Doucet et al.: Neuromuscular Electrical Stimulation for Skeletal Muscle Function
`
`207
`
`human movement, the smaller, fatigue-re-
`sistant motor units are activated first, which
`helps to delay the onset of fatigue; however,
`motor unit recruitment in electrically evoked
`contractions is suggested to be more ran-
`dom, thereby compromising the natural rate
`of fatigue resistance [69]. Although the re-
`versal of Hennemann’s size principle (where
`smaller motor units are recruited before
`larger motor units during voluntary contrac-
`tions) [70] is a commonly reported short-
`coming of NMES; some have postulated
`that, rather than an exact reversal of the
`process, activation may be less systematic
`or non-selective [71]. Jubeau et al. [72] re-
`ported that the when the quadriceps muscle
`belly in 16 healthy men was stimulated with
`NMES, motor units were recruited in a
`“nonselective/random order” regardless of
`fiber type. Additionally, recent work using
`NMES applied over the tibial nerve as com-
`pared to the triceps surae muscle belly ob-
`served that contractions were more forceful,
`activated spinal neurons for increased cen-
`tral nervous system input, and tended to fol-
`low
`the normal physiological motor
`recruitment size principle [73]. Other work
`by Thomas et al. [74] with spinal injured in-
`dividuals indicated that a motor recruitment
`order similar to that which occurs in volun-
`tary muscle contractions could be seen in the
`thenar muscles of the hand when using
`NMES.
`Second, muscle fibers being stimulated
`are done so simultaneously, much unlike the
`normal, unsynchronized, highly-effective re-
`cruitment and derecruitment process of
`motor units seen during voluntary muscle
`contractions. In these contractions, the
`human motor system offsets fatigue by in-
`creasing the firing rate of active motor units
`and/or recruiting new motor units to replace
`others that have been derecruited due to fa-
`tigue [75]. This simultaneous activation ob-
`served during NMES can produce sudden,
`sometimes uncoordinated, inefficient move-
`ment patterns rather than the smooth grada-
`tion of force typically seen in human
`movement.
`Third, surface-stimulating electrodes
`direct current precisely beneath the surface
`
`area of the electrode, and because the cur-
`rent will travel through various viscosities
`of subcutaneous tissue that create resistance,
`its strength will be diminished and the depth
`of penetration will be limited. Fuglevand et
`al. [76] noted that surface-stimulating elec-
`trodes typically reach superficial motor units
`10-12 mm in close proximity to the elec-
`trode face and that only the larger motor
`units are detected from deeper tissues.
`Therefore, activation of deeper structures is
`usually not possible with standard surface
`stimulation; however, increasing pulse width
`or amplitude can improve penetration of cur-
`rent in an effort to reach muscles distant
`from the skin surface [26,77].
`Another limitation of ES is related to its
`questionable long-term effectiveness fol-
`lowing discontinuation. Few studies have
`follow-up data after treatment; however,
`some reports of received benefits waning
`following withdrawal of ES are present
`across different types of applications, such
`as spasticity reduction in children with cere-
`bral palsy [78], functional hand use after-
`stroke [79,80], and shoulder subluxation
`[81]. Therefore, NMES may not be a long-
`term intervention for muscle re-education or
`restoration of movement. However, for SCI,
`some have suggested that only long-term
`use of ES helps to offset the muscle atrophy
`and complications of disuse [82].
`
`VariationS oF ElEctrical
`StiMulation dEliVEry
`
`Another type of transcutaneous stimu-
`lation is electromyography (EMG)-triggered
`electrical stimulation. This type of stimula-
`tion assists patients who are relearning spe-
`cific muscle movements for function.
`Muscle activity is monitored by means of
`EMG recording electrodes such that when
`the EMG signal reaches a specific threshold
`(usually set by therapist), the stimulation
`will activate, thus assisting the patient to
`complete a movement. This intervention has
`been described as being even more reinforc-
`ing than cyclic stimulation due to the pro-
`prioceptive
`feedback
`and voluntary
`component involved [83]. Motor improve-
`
`LUMENIS EX1016
`Page 7
`
`

`

`208 Doucet et al.: Neuromuscular Electrical Stimulation for Skeletal Muscle Function
`
`ments in hand function [84,85] and lower
`extremity motor skills for ambulation [86]
`following stroke have been observed. EMG-
`triggered electrical stimulation has also im-
`proved gait in patients with incomplete
`spinal injury [87].
`Percutaneous stimulation uses electrodes
`that are inserted through the skin into the
`muscle of choice and are thought to be a su-
`perior choice to transcutaneous surface elec-
`trodes when specificity of stimulation is
`paramount. The leads of the electrodes exit
`the skin and connect to an external stimulator,
`bypassing sensory therefore minimizing dis-
`comfort. These hair-thin electrodes can usu-
`ally target specific deeper muscle locations
`without the consequence of unintentionally
`activating surrounding tissues, as often hap-
`pens in transcutaneous applications. The elec-
`trodes can be left in place on average for
`about 3 months, but skin irritation and break-
`ing or dislodging of the electrode can occur
`[61]. Percutaneous FES implants have been
`shown to be effective for significantly reduc-
`ing shoulder pain associated with post-stroke
`glenohumeral subluxation [88,89].
`More recently, small stimulators can be
`surgically implanted for FES applications.
`This is a long-term alternative for stimula-
`tion protocols that require use for extensive
`periods. One of the earliest systems that be-
`came popular for spinal injured persons was
`the NeuroControl Freehand system (Neuro-
`Control, Cleveland, OH). This product con-
`sisted of an implanted stimulator, electrodes,
`and position sensor placed near the shoulder
`joint of the spinal injured individual. The
`system was attached to an external control
`unit for activation. The patient used intact
`shoulder muscles to trigger stimulation to
`paralyzed upper extremity muscles to pro-
`duce a functional grasp and release of the
`dominant hand. In a multi-site randomized
`trial, 49 of 50 patients made improvements
`in grasp, pinch, and functional use of the
`hand, which was maintained 3 years follow-
`ing the implantation [90]. However, due to
`complicating logistical and marketing is-
`sues, the product is no longer available.
`Implanted electrodes also have been
`used to activate spinal nerves to alleviate
`
`back pain or intractable pain associated with
`complex regional pain syndrome; however,
`while initial studies indicate effectiveness,
`extensive evidence for effectiveness is lack-
`ing [91].
`Deep brain stimulation systems im-
`planted directly into cortex are developing
`as a means to decrease symptoms of Parkin-
`son's Disease [92] as well as to control
`seizures in persons with neurological pathol-
`ogy or epilepsy [93].
`
`StiMulation SyStEMS
`currEntly on MarkEt
`
`By far, the most convenient way to
`apply ES is through the small portable units.
`These units have modifiable capabilities so
`therapists can set parameters and design cus-
`tom ES programs that patients can use in the
`clinic or at home. Many come with pre-pro-
`grammed regimens from which the therapist
`can choose that have fixed parameter set-
`tings, depending on the goal of treatment
`(strengthening, muscle re-education, pain re-
`lief, etc.). Most of these units can be locked
`so that patients can take them home without
`fear of altering the program or parameter set-
`tings, and the patient need only turn the unit
`on to activate the set program. Other options
`available on the units are tracking or com-
`pliance mechanisms that monitor activity in
`the unit. This allows the therapist to check
`how often and for what duration the unit was
`turned on, so that compliance with an ES
`program can be determined. Companies cur-
`rently offering small portable units for pa-
`tient use are numerous. Examples of these
`products are the Empi 300 PV (Empi,
`Inc.,www.empi.com),
`a multi-function
`portable device with TENS, NMES, and
`high-voltage stimulation capabilities [94];
`the Chattanooga group (Chattanooga, Inc.,
`www.chattgroup.com) offers portable and
`desktop clinical units with multiple ES op-
`tions as well.
`The Parastep I (Sigmedics,Inc., www.
`sigmedics.com) was one of the first FES am-
`bulatory systems to be approved by the FDA
`and uses an array of stimulation across the
`back, gluteals, and lower extremities. The
`
`LUMENIS EX1016
`Page 8
`
`

`

`Doucet et al.: Neuromuscular Electrical Stimulation for Skeletal Muscle Function
`
`209
`
`Parastep also uses a walker apparatus with
`hand controls to regulate standing and sit-
`ting. Mushahwar et al. [95] summarized that
`Parastep I has modest success in restoring
`upright stance and gait as an activity of daily
`living and is better suited for users with
`complete SCI at the le