Journal of Electromyography and Kinesiology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Journal of Electromyography and Kinesiology journal homepage: www.elsevier.com/locate/jelekin

Review

Advances in functional electrical stimulation (FES) Dejan B. Popovic´ ⇑ University of Belgrade, Faculty of Electrical Engineering, Belgrade, Serbia Serbian Academy of Sciences and Arts (SASA), Belgrade, Serbia

a r t i c l e

i n f o

a b s t r a c t

Article history: Received 30 August 2014 Accepted 9 September 2014 Available online xxxx

This review discusses the advancements that are needed to enhance the effects of electrical stimulation for restoring or assisting movement in humans with an injury/disease of the central nervous system. A complex model of the effects of electrical stimulation of peripheral systems is presented. The model indicates that both the motor and sensory systems are activated by electrical stimulation. We propose that a hierarchical hybrid controller may be suitable for functional electrical stimulation (FES) because this type of controller acts as a structural mimetic of its biological counterpart. Specific attention is given to the neural systems at the periphery with respect to the required electrodes and stimulators. Furthermore, we note that FES with surface electrodes is preferred for the therapy, although there is a definite advantage associated with implantable technology for life-long use. The last section of the review discusses the potential need to combine FES and robotic systems to provide assistance in some cases. Ó 2014 Elsevier Ltd. All rights reserved.

Keywords: Functional electrical stimulation Mimetic model Neurorehabilitation Stimulators Electrodes

Contents 1. 2. 3. 4. 5. 6.

Introduction . . . . . . . . . . . . Control methods for FES . . Electrodes and stimulators Sensors for FES systems . . . Hybrid systems. . . . . . . . . . Message to take home . . . . Conflict of interest . . . . . . . Acknowledgments . . . . . . . References . . . . . . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

1. Introduction Damage to the central nervous system (CNS) due to injury or disease, in conjunction with other health problems (e.g., muscle atrophy, joint contractures, increased frequency of bladder infections, decreased cardio-vascular capacity), can lead to decreased sensory–motor performance. The extent and type of injury/disease determines the regions of the body that are affected (e.g., cervical lesions resulting in tetraplegia, thoracic lesions resulting in paraplegia, brain lesions resulting in hemiplegia or cerebral palsy). A ⇑ Address: Faculty of Electrical Engineering, Bulevar kralja Aleksandra 73, 11000 Belgrade, Serbia. Tel.: +381 113218348; fax: +381 113248681. E-mail address: [email protected]

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

00 00 00 00 00 00 00 00 00

sensory–motor disability directly affects the patient’s lifestyle and limit life activities. Functional electrical stimulation (FES) was introduced as a method to artificially activate the sensory motor system after a CNS injury/disease and alleviate the resulting disability. FES systems are often used in motor neural prostheses (e.g., the peroneal stimulator for stroke patients (Burridge et al., 2007; Everaert et al., 2013) and may be designed as surface or implantable stimulation systems for control of the arm and hand (Hart et al., 1988; Popovic´ et al., 2004) or for control of standing and walking (Kralj and Bajd, 1989; Nataraj et al., 2012; Dutta et al., 2011; Popovic´ et al., 2003)). FES devices directly assist in the performance of disrupted functions in humans with CNS lesions; furthermore, FES causes changes in cortical excitability and stimulates cortical

http://dx.doi.org/10.1016/j.jelekin.2014.09.008 1050-6411/Ó 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Popovic´ DB. Advances in functional electrical stimulation (FES). J Electromyogr Kinesiol (2014), http://dx.doi.org/ 10.1016/j.jelekin.2014.09.008

2

D.B. Popovic´ / Journal of Electromyography and Kinesiology xxx (2014) xxx–xxx

reorganization (carry-over effects). Electrical stimulation can act directly on the central nervous system (Visser-Vandewalle et al., 2004; van den Brand et al., 2012), or it can be applied to the peripheral nervous system (Fisher et al., 2008; Popovic´ and Sinkjær, 2000; Popovic´ et al., 2009). The majority of studies reported in the literature describe sophisticated FES systems that can be used for the restoration of sensory–motor function; however, to date, only simulation studies or anecdotal data from healthy individuals or case studies have been reported for these systems. By contrast, the systems that are currently in use and have been tested in patients are rather simple (reviewed in Popovic´ et al. (2009), Popovic´ and Sinkjær (2000)). During FES, bursts of short pulses of electrical charge generate an electrical field that triggers action potentials in afferent and efferent neural pathways. The externally triggered efferent pathways directly activate muscles that are innervated by neurons, but this activation differs from activation by a volitional motor command from an upper motor neuron. In parallel, the activity triggered in afferent pathways carries action potentials to the spinal cord where various reflexes are generated (e.g., the crossextension reflex and the flexion reflex), and interneurons are activated and transmit signals that eventually reach the cortex (Fig. 1).

This model of FES is highly complex compared with the commonly assumed models in which only the motor component is considered and the effects of the upper motor neuron are not addressed. Typical modeling strategies consider the muscles as the actuators, the joint trajectory as the input and the electrical stimulation as the output. The complexity of these models arises from the number of joints included, not from consideration of the biological network involved in sensory–motor control. The model presented here is simplified with respect to the complexity of a living system because the externally generated activity of the targeted muscle also results in a change in the activity of antagonistic muscles and may also affect environmental interactions. In this model, the neural traffic is augmented by proprioception and exteroception signals which follows externally activated body segments. This integration of neural systems was demonstrated in targeted reinnervation applied in arm prosthetics (Kuiken et al., 2007), and it is possible that this approach may eventually be used in FES systems. In this paper, we discuss technological developments in the domains of miniaturized stimulators, electrodes, sensors and wireless communication as well as the results of clinical trials and the commercialization of FES systems.

Fig. 1. Model of the effects of peripheral electrical stimulation. The stimulator generates a voltage that creates a pulsatile electrical field that activates afferent and efferent neurons, resulting in direct activation of the muscles that are innervated by the stimulated nerve and several reflex responses. The top part of the figure presents two aspects that need to be considered: the lesion and the complex networking involved in the control of movement.

Please cite this article in press as: Popovic´ DB. Advances in functional electrical stimulation (FES). J Electromyogr Kinesiol (2014), http://dx.doi.org/ 10.1016/j.jelekin.2014.09.008

D.B. Popovic´ / Journal of Electromyography and Kinesiology xxx (2014) xxx–xxx

3

Fig. 2. A model of a hybrid hierarchical controller for functional electrical stimulation (right panel) in parallel with the simplified model of natural control (left panel).

2. Control methods for FES The mimesis of biological control has the highest likeliness for the successful restoration of function (van den Brand et al., 2012; Popovic´ and Popovic´, 2011). The mimetic model has a hierarchical hybrid structure (HHS) (Tomovic´ et al., 1995), as shown in the right panel of Fig. 2. The highest levels consider the finite states (discrete control), while the lower levels include dynamics (continuous, model-based control). The adjective ‘‘hybrid’’ describes behavior that can be defined by the interaction of subsystems in the context of both continuous dynamics and discrete events. The hierarchical organization of the controller facilitates the management of its complexity. The higher levels of the hierarchy require less detailed models (discrete abstractions) of the functioning at the lower levels. The lower levels incorporate the interactions of discrete and continuous components (Tomovic´ et al., 1995). The block ‘‘Sensors’’ (intention detection) in Fig 2 represent the interface between the user and the FES. This communication channel allows the user to trigger/select the operation and to adapt the process based on the sensors that provide on-line information about the effects. The brain–computer interface (BCI) is a promising type of sensor that detects the activity of the biological system (Savic´ et al., 2014). Examples of the effective implementation of a BCI for the control of FES are presented elsewhere (Pfurtscheller et al., 2003, 2005). The BCI is gaining popularity because it is directly controlled, possibly at the subconscious level, by the user. The BCI is also attracting attention due to important developments in the miniaturization of devices for recording brain activity (e.g., http://www.mBrainTrain.com) and advanced processing techniques. Alternatively, a rather invasive method for interfacing with the brain is to use a brain–machine interface (BMI), which provides much more data compared with the BCI (e.g., Schwartz et al., 2006). Artificial perception is another new modality for the recognition of intention. In this case, artificial vision is controlled by the user and provides signals that are sufficient for the control of an assistive device (Došen et al., 2010; Markovic´ et al., 2014; Hao et al., 2013; Štrbac et al., 2014). Once the user’s intention is detected, a feedback system using discrete and sampled data selects a program from a database created from observations and synergistic models of movement

(Grasso et al., 2004; Ivanenko et al., 2004). This method of control is equivalent to biological control at the level of the brainstem and spinal cord. The discrete models include a central motor pattern generator (Dietz, 2003) if required (e.g., cyclic activities, such as the gait or cycling). The discrete model incorporates two elements: temporal and spatial synergies that handle the timing and the individual activities at the joint level (Popovic, 2003). A well-suited method for the implementation of discrete control is known as finite state control (FSC). FSC is a symbolic technique that relies on non-parametric models of movement (Tomovic´ et al., 1995). Non-parametric models use set theory and symbols in a multidimensional phase space. FSC is suitable because it addresses the redundancy, nonlinearity and time variability of the system. FSC operates based on three components: (1) the rule base, which consists of the set of production rules; (2) data structures containing the known facts relevant to the domain of interest; and (3) the interpreter of these facts and rules. An important feature of an FSC system is the ability to look first at the established facts and to proceed forward (forward chaining) or to start from the aims, i.e., from the action components of the rules (backward chaining). The rules in FSC can be either heuristically defined through a procedure known as ‘‘hand-crafting’’, or they can be automatically generated thorough machine classification (Tomovic´ et al., 1995; Došen and Popovic, 2008; Jonic´ et al., 1999; Kostov et al., 1995; Nikolic and Popovic, 1998). Continuous feedback at the joint level is essential for smooth, biological-like movement because of the properties of the musculo-skeletal system. The model must incorporate the properties of the sensory–motor systems in a given subject because an impairment will result in differences compared with a healthy subject, and sensory–motor systems vary greatly between individuals with disabilities. This level of the model is responsible for the activation of specific muscle groups and is essential for the provision of functional movements. The model-based control relies on a skeletal model of the body with the joints driven by externally activated muscles. The model must include account for the tendons and ligaments that connect muscles and bones and control the stiffness of the joints. One important aspect that remains to be resolved is the time variability of the responses of muscles (e.g., muscle fatigue, habituation, etc.). In addition, it is not easy to incorporate bi- and multi-articular muscles into the model. It is even more difficult

Please cite this article in press as: Popovic´ DB. Advances in functional electrical stimulation (FES). J Electromyogr Kinesiol (2014), http://dx.doi.org/ 10.1016/j.jelekin.2014.09.008

D.B. Popovic´ / Journal of Electromyography and Kinesiology xxx (2014) xxx–xxx

4

Fig. 3. A model of a joint based on the physiological properties of stimulated agonistic muscles and passive antagonistic muscle. The model basically considers all of the muscles acting on a joint as a single (equivalent) muscle. The model must be adapted if bi-articular and multi-articular muscles are considered for activation. The complete model should include tendons and ligaments.

to include the reflex responses of antagonistic muscles when the action is generated in the agonist muscle. A model of the joint derived from detailed studies (e.g., Shue et al., 1992) is shown in Fig. 3. The nonlinear properties of a muscle (muscle force vs. length, muscle force vs. the velocity of shortening, recruitment and activation dynamics) are specific for each muscle (e.g., Stein et al., 1999). Estimation of the skeletal parameters and the corresponding ligament and tendon properties has been partially resolved with sufficient accuracy in humans with disabilities (e.g., Chizeck et al., 1999). These findings suggest that model-based control is an efficient and important method for simulating the behavior of a system; however, this approach is not suitable for real-time control.

3. Electrodes and stimulators Fig. 1 shows a model of a biological system that is activated by external electrical stimulation at the periphery. The interface between the source of the electrical charge (stimulator) and the

tissues is provided by electrodes. The electrodes can be applied to the skin or implanted, and they can either be connected to an external stimulator with wires or constructed as part of the implant. The electrodes must be biocompatible, safe, have an appropriate longevity, and provide sufficient selectivity to guaranty that only the targeted motor system are activated. Here, we address only the selectivity problem, assuming that issues related to safety and the long-term operation of interface materials have been resolved. To document the complexity of the selectivity, we first discuss the anatomy and physiology of the peripheral systems of the arm/hand (Fig. 4). The ulnar nerve enters the anterior compartment of the forearm through the two heads of the flexor carpi ulnaris and runs alongside the ulna. The ulnar nerve travels deep to the flexor carpi ulnaris and divides into three branches: (1) the muscular branches of the ulnar nerve, (2) the palmar branch of the ulnar nerve, and (3) the dorsal branch of ulnar nerve. The ulnar nerve enters the palm of the hand by passing superficial to the flexor retinaculum via the ulnar canal and divides into superficial and deep branches.

Fig. 4. Cross section of the forearm, major nerves of the arm and hand (middle panel), and the brachial plexus annotation of the spinal nerves (right panel).

Please cite this article in press as: Popovic´ DB. Advances in functional electrical stimulation (FES). J Electromyogr Kinesiol (2014), http://dx.doi.org/ 10.1016/j.jelekin.2014.09.008

D.B. Popovic´ / Journal of Electromyography and Kinesiology xxx (2014) xxx–xxx

5

Fig. 5. Images of the peripheral nerves of two arms in cadavers (top and bottom panels).

The ulnar nerve innervates many muscles (e.g., the flexor carpi ulnaris and flexor digitorum profundus (medial half), the hypothenar, opponens digiti minimi, abductor digiti minimi, flexor digiti minimi brevis, and adductor pollicis, the third and fourth lumbricals, the dorsal interossei, and the palmar interossei). Through the superficial branch, it innervates the palmaris brevis. The ulnar nerve transmits sensory information from the cutaneous nerve and other nerves of the fourth and fifth digits. The radial nerve originates from the posterior cord of the brachial plexus. This nerve innervates many muscles and transmits afferent signals from most of the back of the hand. The median nerve is formed from parts of the medial and lateral cords of the brachial plexus and innervates many muscles. The median nerve afferent function is related to the lateral part of the palm, the skin of the palmar side of the thumb, the index and middle finger, half of the ring finger, and the nail beds of these fingers. Based on the descriptions above, it is clear that even if only one of the three listed nerves is activated completely, many agonist and antagonistic muscles would contract and it would be very difficult to control the level of force. Therefore, electrodes should target individual motor units directly and not whole nerves. The second issue to be considered relates to the nerve topology, which differs from subject to subject and must be individually set for effective application of FES. Namely, the nerve endings at their sites of connection to the muscles are different between individuals. These differences are evident in images (Fig. 5) acquired during

Fig. 6. Nerve diameter and the velocity of propagation of action potentials for the sensory system in the arm. Merkel, Ruffini, Crouse, and Pacinian cells and hair receptors are exteroceptive sensors, while the muscle spindles and Golgi tendon organs are proprioceptive sensors.

Fig. 7. The discharge rate for one and two adjunct points on the skin. Modified from Textbook of Medical Physiology, A.C. Guyton, Saunders, 1991.

our cadaver studies (unpublished results acquired with AJ Hoffer at the University of British Columbia, Vancouver, Canada). This work followed strict ethical principles approved by the NIH, Bethesda, USA, and the local ethics committee. The third aspect to be considered is the threshold for stimulation based on the size of the specific nerves. Fig. 6 shows the diameter and the velocity of propagation of action potentials of the sensory fibers in the arm. It is necessary to set different intensities of stimulation based on the diameter of the neural pathway that is being targeted because the threshold for generating an action potential varies with the size of the nerve. The same is true for the motor nerves. Finally, it is important to consider the neurophysiology related to the firing rate (Fig. 7). The direct consequence of the fact that the discharge rate is correlated with the strength of the stimulation can be used as a strategic element for activating peripheral nerves. Advanced FES systems must provide fine control through variation of the amplitude, pulse duration and firing rate for each stimulation channel in systems with implantable or surface array electrodes. Stimulators are electronic components of the FES that ultimately integrate the source of a train of pulses of electrical charge by elevating the low voltage provided by a battery to a high

Please cite this article in press as: Popovic´ DB. Advances in functional electrical stimulation (FES). J Electromyogr Kinesiol (2014), http://dx.doi.org/ 10.1016/j.jelekin.2014.09.008

6

D.B. Popovic´ / Journal of Electromyography and Kinesiology xxx (2014) xxx–xxx

voltage by means of a DC/DC converter with an output stage. This device controls the current in the form of biphasic, charge compensated pulses within the range determined by the type of electrodes used. The rate of pulses, their individual duration and amplitude, the type of charge compensation, and the rise and fall time are controlled by a microcomputer that is part of the stimulator. The advanced stimulators use high efficiency DC/DC converters, lowpower fast microcontrollers with multiple analog-to-digital converters and multiple digital outputs controlling the output stages. Stimulators for implantable systems benefit from developments in the pacemaker industry (packaging, battery source, biocompatible materials). Most advanced stimulators have wireless communication with Windows or Android platforms for off-line setup (e.g., Maleševic´ et al., 2012). The greatest challenges in the domain of implantable stimulators are related to the connection of the electrodes, the low capacity of the batteries, the difficulties associated with recharging a device that is inside the body, and communication with external sensors required for sensor-driven or closed-loop control. With respect to the electrodes, the best choices are implants that interface directly with neurons. Cuff electrodes introduced in the 1970s are now available and functional with an array of contact points (Durand et al., 2005; Rodriguez et al., 2000). These electrodes can direct the electrical field within the nerve and provide the necessary selective response. The alternative is to use intraneural electrodes, such as transversal intrafascicular electrodes (Boretius et al., 2010; Raspopovic et al., 2014), electrodes such as the Utah slant array (Branner et al., 2001) or electrode arrays that can be positioned along the spinal cord (Gad et al., 2013). Based on the discussion presented above, the following conclusions can be stated: (1) Array multi-pad electrodes can be implanted in or around the peripheral nerves (Navarro et al., 2005). There are several possible sites where the nerves can be separated from the surrounding tissues at lengths that are sufficient to house cuff electrodes or intraneural electrodes; (2) By varying the potentials of the electrodes that are in contact with the neural tissues, it is possible to steer the currents; thereby, selectively activating portions of the nerve; (3) By varying the stimulation rate, it is possible to activate peripheral nerves in a manner that will produce different effects on the muscles and thereby generate distinct episodes that are strongly correlated with the strength of the stimulation; and (4) By varying the level of stimulation, it is possible to activate motor and proprioceptive fibers, generate muscle contraction and provide information about the joint angles and joint angular velocities (muscle spindles and Golgi tendon organ replacement) or activate smaller nerves and substitute for the lack of exteroception. However, the implantable electrodes of this type are not suitable for therapy. An alternative is the use of intramuscular electrodes that are implanted by a needle and removed after the therapy by simply pulling at their free end (Daly et al., 2000). The BION implantable system has been suggested and tested in several applications (Loeb et al., 2001). BION is a miniature stimulation unit that integrates the electrodes and communicates with the central control unit and receives energy wirelessly. A later version of BION includes a rechargeable battery, but this resulted in the size of the device being too large to be suitable for implantation by a subdermal needle as originally suggested. A new solution for sending the electrical current selectively deep into the tissues is the so-called ‘‘router’’ system in which the electrodes are implanted, but they are activated wirelessly through the skin (Gan et al., 2007). For therapy, the use of surface electrodes is greatly preferred because they are not invasive and the price range of their application is appropriate. Currently, surface electrodes are applied using a biocompatible gel that provides stabilization on the skin and

excellent distribution of the current over the surface of the electrode. Conventional surface electrodes are suitable for large muscles with innervation close to the skin. To improve the selectivity of FES for activities such as extending the fingers while not activating the wrist extensors or flexing fingers while not activating wrist flexors, the use of a new electrode array is preferred (Popovic´ and Popovic´, 2009). Available technology that automatically selects and asynchronously activates some pads of the electrode array (Maleševic´ et al., 2012) provides the appropriate control of functions (Popovic´-Maneski et al., 2013a). Asynchronous stimulation is also a key for reducing the stimulation rate, thereby postponing the onset of muscle fatigue (Popovic´-Maneski et al., 2013b). This important finding also supports the use of electrode arrays on large muscles (e.g., the quadriceps while assisting in standing/walking) because the time before the onset of fatigue can be greatly increased (Maleševic´ et al., 2010). The current technology allows the use of basically any shape of electrode with many small pads (area approximately 1 cm2) and the conductive gel. 4. Sensors for FES systems Micro- and nano-technology allow real-time monitoring of position, velocity, acceleration, orientation in space, distance, and many other physical properties. The current sensors are small, require low energy and can be integrated with a microcomputer and wireless communication circuitry. Although the sensors are miniature, they are still not acceptable for certain applications (e.g., fingers and hand). At the moment, there is no system that allows for non-invasive measurement of the force generated by an individual muscle. An estimate of this force could be derived from electromyography (EMG) signals. Assessment of muscle activity from the EMG is rather difficult in the presence of electrical stimulation, particularly if more stimulation channels are applied via electrodes that are close to each other. Blanking of the stimulation artifact is feasible, but this approach has not yet been perfected (Frigo et al., 2000; Thorsen and Ferrarin, 2009). The assessment of the level of contractions using ultrasound or sound is not feasible with the available technologies. The assessment of ground reactions is receiving considerable attention, but there are still major challenges for obtaining a sufficiently accurate and practical real-time estimation of the distribution of the force over the sole while walking. The issues are mostly related to the comfort of walking while testing sophisticated sensor systems and the uncertainty of measurements obtained with the more comfortable interfaces. It is important to note that if used in closed loop control configurations, sensors need to operate in real time, and even then, they are not appropriate due to the delays in characterizing the muscle response. The sensors used in sensor-driven control systems are appropriate, but the reliability and measurements uncertainty of these systems remains an issue. The use of biological sensors is an attractive idea (Sinkjær et al., 2003). Biological sensors communicate with the central nervous system through afferent neurons; therefore, the information can be extracted from recordings of peripheral nerves (Popovic´ et al., 1993). 5. Hybrid systems FES can be applied and will result in muscle contraction if the muscles are innervated. Stimulation of denervated muscles with a large electrical charge will generate contraction, but the force of the contraction will remain below the useful range for movement or for the control of joint stiffness. In general, the muscle out-

Please cite this article in press as: Popovic´ DB. Advances in functional electrical stimulation (FES). J Electromyogr Kinesiol (2014), http://dx.doi.org/ 10.1016/j.jelekin.2014.09.008

D.B. Popovic´ / Journal of Electromyography and Kinesiology xxx (2014) xxx–xxx

put produced by electrical stimulation in persons with disability is lower compared to the output characteristic for healthy individuals. The impairment also leads to atrophy and may result in joint contractures and changes in the lengths of the unused muscles (e.g., the range of movement during simultaneous extension of the hip and knee could be reduced). The solution for situations in which FES is not appropriate is the combination of a light, modular exoskeleton that will provide the needed functions, as originally suggested by Popovic´ et al. (1989). The implementation of a hybrid system that combined reciprocal gait orthosis and four channels of electrical stimulation was described by Solomonow et al. (1989). The use of hybrid systems is gaining attention because many robotic systems have been introduced (Ferguson et al., 1999; Kobetic et al., 2009). 6. Message to take home Functional electrical stimulation is a powerful tool for inducing the contraction of paralyzed muscles. The mobility of complete paraplegics and full functionality for object manipulation and grasping in complete tetraplegics remain to be achieved. However, the carry-over effects suggested in the literature in post-stroke patients and patients with incomplete spinal cord injuries are promising. The therapeutic use of FES is benefiting from the new array electrodes and effective stimulators, particularly when integrated with exercises to provide excitation feedback. The therapeutic use of FES is also improving with the use of the BCI for control. An important element that needs to be improved is the ease of mounting and the use of surface stimulation systems, which would allow users and clinicians to focus their attention on the therapy itself and not on how to operate the system. The implantable systems have a definite advantage when a prolonged period of use is required. The main challenges are related to the technology used for the electrodes, connectors, and energy supply for the stimulator. The electrodes must be designed in a manner that will not harm the preserved sensory–motor systems and will not deteriorate due to environmental conditions. The connectors should be safe, small and allow eventual replacement of the electrode or the stimulator. The power supply for the stimulator should be rechargeable and also reprogrammable. Ultimately, this objective requires efficient energy transmission from and to the stimulator. Neither surface nor implantable systems provide adequate feedback to the user. The sensors need to be integrated into a network that provides data similar to a natural sensing network. In this domain, it is even more important that neuroscience research and clinical testing provide solid evidence regarding the optimal means of feedback. The feedback must allow the use of the system at the subconscious level and support its integration with the preserved biological control mechanism. Sensor development is equally important for the use of FES and the feedback provided to the user. Finally, through clinical studies, researchers need to demonstrate to caregivers and healthcare providers the benefits of FES that make these systems commercially feasible for users and clinicians. Conflict of interest The author declares that there are no conflict of interest. Acknowledgments The research that led to this review paper was partly supported by the grant from the Ministry of Education, Science and Techno-

7

logical Development, Republic of Serbia, No. 175016 and the Danish National Research Foundation.

References Boretius T, Badia J, Pascual-Font J, Schuettler M, Navarro X, Yoshida K, et al. A transverse intrafascicular multichannel electrode (TIME) to interface with the peripheral nerve. Biosens Bioelectron 2010;26(1):62–9. Branner A, Stein RB, Normann RA. Selective stimulation of cat sciatic nerve using an array of varying-length microelectrodes. J Neurophysiol 2001;85:1585–94. Burridge JH, Haugland M, Larsen B, Pickering RM, Svaneborg N, Iversen HK, et al. Phase II trial to evaluate the ActiGait implanted drop-foot stimulator in established hemiplegia. J Rehabil Med 2007;39:212–8. Chizeck HJ, Chang S, Stein RB, Scheiner A, Ferencz DC. Identification of electrically stimulated quadriceps muscles in paraplegic subjects. IEEE Trans Biomed Eng 1999;46(1):51–61. Daly JJ, Ruff RL, Haycook R, Strasshofer B, Marsolais EB, Dobos L. Feasibility of gait training for acute stroke patients using FNS with implanted electrodes. J Neurol Sci 2000;179(1–2):103–7. Dietz V. Spinal cord pattern generators for locomotion. Clin Neurophysiol 2003;114(8):1379–89. Došen S, Popovic´ DB. Accelerometers and force sensing resistors for optimal control of walking of a hemiplegic. IEEE Trans Biomed Eng 2008;55(8):1973–84. Došen S, Cipriani C, Kostic´ M, Carrozza MC, Popovic´ DB. Cognitive vision system for the control of a dexterous prosthetic hand: an evaluation study. J Neuroeng Rehabil 2010;7:42. http://dx.doi.org/10.1186/1743-0003-7-42. Durand DM, Grill WM, Kirsch R. Electrical stimulation of the neuromuscular system. In: He B, editor. Neural engineering. New York: Springer-Verlag; 2005. p. 157–91. Dutta A, Kobetic R, Triolo RJ. An objective method for selecting command sources for myoelectrically triggered lower-limb neuroprostheses. J Rehabil Res Dev 2011;48:935–48. Everaert DG, Stein RB, Abrams GM, DromerickH AW, Francisco GE, Hafner BJ, et al. Effect of a foot-drop stimulator and ankle-foot orthosis on walking performance after stroke: a multicenter randomized controlled trial. Neurorehabil Neural Repair 2013;27(7):579–91. Ferguson KA, Polando G, Kobetic R, Triolo RJ, Marsolais EB. Walking with a hybrid orthosis system. Spinal Cord 1999;37(11):800–4. Fisher LE, Miller ME, Bailey SN, Davis Jr HA, Anderson JS, Murray LR, et al. Standing after spinal cord injury with four-contact nerve-cuff electrodes for quadriceps stimulation. IEEE Trans Neural Syst Rehabil Eng 2008;16:473–8. Frigo C, Ferrarin M, Frasson W, Pavan E, Thorsen R. EMG signals detection and processing for on-line control of functional electrical stimulation. J Electromyog Kinesiol 2000;10(5):351–60. Gad P, Choe J, Nandra MS, Zhong H, Roy RR, Tai YT, et al. Development of a multielectrode array for spinal cord epidural stimulation to facilitate stepping and standing after a complete spinal cord injury in adult rats. J Neuroeng Rehabil 2013;10:2. Gan LS, Prochazka A, Bornes TD, Denington AA, Chan KM. A new means of transcutaneous coupling for neural prostheses. IEEE Trans Biomed Eng 2007;54(3):509–17. Grasso R, Ivanenko RP, Zago M, Molinari M, Scivoletto G, Castellano V, et al. Distributed plasticity of locomotor pattern generators in spinal cord injured patients. Brain 2004;127(5):1019–34. Hao Y, Controzzi M, Cipriani C, Popovic´ DB, Yang X, Chen W, et al. Electrooculography as a substitute of vision for control of prehension of hand assistive devices. IEEE Robot Automat Mag 2013;20(1):40–52. Hart RL, Kilgore KL, Peckham PH. A comparison between control methods for implanted FES hand-grasp systems. IEEE Trans Rehabil Eng 1988;6:208–18. Ivanenko YP, Poppele RP, Lacquaniti F. Five basic muscle activation patterns account for muscle activity during human locomotion. J Physiol 2004;556:267–82. Jonic´ S, Jankovic T, Gajic V, Popovic´ DB. Three machine learning techniques for automatic determination of rules to control locomotion. IEEE Trans Biomed Eng 1999;46(3):300–10. Kobetic R, To CS, Schnellenberger JR, Audu ML, Bulea TC, Gaudio R, et al. Development of hybrid orthosis for standing, walking, and stair climbing after spinal cord injury. J Rehabil Res Dev 2009;46(3):447–62. Kostov A, Andrews B, Popovic´ DB, Stein RB, Armstrong WW. Machine learning in control of functional electrical stimulation (FES) for locomotion. IEEE Trans Biomed Eng 1995;42:541–51. Kralj A, Bajd T. Functional electrical stimulation: standing and walking after spinal cord injury. CRC Press; 1989. Kuiken TA, Marasco PD, Lock BA, Harden RN, Dewald JP. Redirection of cutaneous sensation from the hand to the chest skin of human amputees with targeted reinnervation. Proc Natl Acad Sci USA 2007;104(50):20061–6. Loeb GE, Peck RA, Moore WH, Hood K. BION system for distributed neural prosthetic interfaces. Med Eng Phys 2001;23:9–18. Maleševic´ N, Popovic´ L, Schwirtlich L, Popovic´ DB. Distributed low-frequency electrical stimulation delays muscle fatigue compared to conventional stimulation. Muscle Nerve 2010;42:556–62. Maleševic´ NM, Popovic´-Maneski LZ, Ilic´ V, Jorgovanovic´ N, Bijelic´ G, Keller T, et al. A multi-pad electrode based functional electrical stimulation system for restoration of grasp. J Neuro Eng Rehabil 2012;9:66. http://dx.doi.org/ 10.1186/1743-0003-9-66.

Please cite this article in press as: Popovic´ DB. Advances in functional electrical stimulation (FES). J Electromyogr Kinesiol (2014), http://dx.doi.org/ 10.1016/j.jelekin.2014.09.008

8

D.B. Popovic´ / Journal of Electromyography and Kinesiology xxx (2014) xxx–xxx

Markovic´ M, Došen S, Cipriani C, Popovic´ DB, Farina D. Stereovision and augmented reality for closed-loop control of grasping in hand prostheses. J Neural Eng 2014:11. http://dx.doi.org/10.1088/1741-2560/11/4/04600. Nataraj R, Audu ML, Triolo RJ. Comparing joint kinematics and center of mass acceleration as feedback for control of standing balance by functional neuromuscular stimulation. J Neuroeng Rehabil 2012;6:25. http:// www.jneuroengrehab.com/content/9/1/25. Navarro X, Krueger TB, Lago N, Micera S, Stieglitz T, Dario P. A critical review of interfaces with the peripheral nervous system for the control of neuroprostheses and hybrid bionic systems. J Peripher Nerv Syst 2005;10(3):229–58. Nikolic´ Z, Popovic´ DB. Predicting quadriceps muscle activity during gait with an automatic rule determination method. IEEE Trans Biomed Eng 1998;55:1081–5. Pfurtscheller G, Müller GR, Pfurtscheller J, Gerner HJ, Rupp R. ‘‘Thought’’ – control of functional electrical stimulation to restore hand grasp in a patient with tetraplegia. Neurosci Lett 2003;351:33–6. Pfurtscheller G, Müller-Putz GR, Pfurtscheller J, Rupp R. EEG-based asynchronous BCI controls functional electrical stimulation in a tetraplegic patient. EURASIP J Appl Signal Process 2005;19:3152–5. Popovic´ MB. Control of neural prostheses for grasping and reaching. Med Eng Phys 2003;25(1):41–50. Popovic´ DB, Popovic´ MB. Automatic determination of the optimal shape of the surface electrode: selective stimulation. J Neurosci Meth 2009;178(1):174–81. Popovic´ DB, Popovic´ MP. Advances in the use of electrical stimulation for the recovery of motor function. Prog Brain Res 2011;194:215–25. Popovic´ DB, Sinkjær T. Control of movement for the physically disabled. London: Springer; 2000. Popovic´ DB, Tomovic´ R, Schwirtlich L. Hybrid assistive system – neuroprosthesis for motion. IEEE Trans Biomed Eng 1989;37:729–38. Popovic´ DB, Stein RB, Jovanovic´ K, Dai R, Kostov A, Armstrong WW. Sensory nerve recording for closed-loop control to restore motor functions. IEEE Trans Biomed Eng 1993;40:1024–31. Popovic´ DB, Radulovic´ M, Schwirtlich L, Jaukovic´ N. Automatic vs. hand-controlled walking of paraplegics. Med Eng Phys 2003;25:63–74. Popovic´ DB, Popovic´ MB, Sinkjær T, Stefanovic´ A, Schwirtlich L. Therapy of paretic arm in hemiplegic subjects augmented with a neural prosthesis: a cross-over study. Can J Physiol Pharmacol 2004;82:749–56. Popovic´ DB, Sinkjær T, Popovic´ MB. Electrical stimulation as a means for achieving recovery of function in stroke patients. J Neuro Rehabil 2009;25:45–58. Popovic´-Maneski L, Kostic´ M, Keller T, Mitrovic´ S, Konstantinovic´ L, Popovic´ DB. Multi-pad electrode for effective grasping: design. IEEE Trans Neural Syst Rehabil Eng 2013a;21(4):648–54. Popovic´-Maneski L, Maleševic´ N, Savic´ A, Popovic´ DB. Spatially distributed asynchronous stimulation delays muscle fatigue. Muscle Nerve 2013b;48:930–7. Raspopovic S, Capogrosso M, Petrini FM, Bonizzato M, Rigosa J, di Pino G, et al. Restoring natural sensory feedback in real-time bidirectional hand prostheses. Sci Transl Med 2014;6:222ra19. http://dx.doi.org/10.1126/scitranslmed. 3006820. Rodriguez FJ, Ceballos D, Schuttler M, Valero A, Valderrama E, Stieglitz T, et al. Polyimide cuff electrodes for peripheral nerve stimulation. J Neurosci Meth 2000;98(2):105–18.

Savic´ MA, Maleševic´ N, Popovic´ MB. Feasibility of a hybrid brain–computer interface for advanced functional electrical therapy. Sci World J 2014. Article ID 797128, http://dx.doi.org/10.1155/2014/797128. Schwartz AB, Cui T, Weber DJ, Moran DW. Brain-controlled interfaces: movement restoration with neural prosthetics. Neuron 2006;52(1):205–20. Shue G, Crago PE, Chizeck HJ. Muscle-joint models incorporating activation dynamics, moment-angle, and moment-velocity properties. IEEE Trans Biomed Eng 1992;42:212–23. Sinkjær T, Haugland M, Inmann A, Hansen M, Nielsen DK. Biopotentials as command and feedback signals in functional electrical stimulation systems. Med Eng Phys 2003;25(1):29–40. Solomonow M, Baratta R, Hirokawa S, Rightor N, Walker W, Beaudette P, et al. Muscle stimulation powered orthosis as a practical walking system for thoracic paraplegics. Orthopedics 1989;12(10):1309–15. Stein RB, Zehr EP, Lebiedowska MK, Popovic´ DB, Scheiner A, Chizeck HJ. Estimating mechanical parameters of leg segments in individuals with and without physical disabilities. IEEE Trans Rehabil Eng 1999;4:201–11. Štrbac M, Kocˇovic´ S, Markovic´ M, Popovic´ DB. Stereovision and augmented reality for closed-loop control of grasping in hand prostheses. BioMed Res Int (J Biomed Biotechnol) 2014:12. Article ID 740469, http://dx.doi.org/10.1155/2014/ 740469. Thorsen R, Ferrarin M. Battery powered neuromuscular stimulator circuit for use during simultaneous recording of myoelectric signals. Med Eng Phys 2009;31(8):1032–7. Tomovic´ R, Popovic´ DB, Stein RB. Nonanalytical methods for motor control. Singapore: World Science Publisher; 1995. van den Brand R, Heutschi J, Barraud Q, DiGiovanna J, Bartholdi K, Huerlimann M, et al. Restoring voluntary control of locomotion after paralyzing spinal cord injury. Science 2012;336(6085):1182–5. Visser-Vandewalle V, Temel Y, van der Linden Ch, Ackermans L, Beuls E. Deep brain stimulation in movement disorders: the applications reconsidered. Acta Neurol Bel 2004;104:33–6.

Dejan B. Popovic´ Professor of Biomedical Engineering, University of Belgrade, Serbia. He received his PhD in 1982 from the University of Belgrade, Serbia. Member of the Serbian Academy of Sciences and Arts (SASA). He published about 500 publications, out of that about 100 in peer reviewed journals, several books and holds several patents all in the domain of assistive systems for humans with disabilities. He was professor at the University of Alberta, Edmonton, Canada; University of Miami, Florida, USA and Aalborg University, Denmark. He is Editor of the Journal of Automatic Control, Belgrade; associated Editor of the IEEE Transactions on Neural Systems and Rehabilitation Engineering; member of the board of Medical Engineering and Physics and Neurorehabilitation journals. He is fellow and the founding member of the EAMBES.

Please cite this article in press as: Popovic´ DB. Advances in functional electrical stimulation (FES). J Electromyogr Kinesiol (2014), http://dx.doi.org/ 10.1016/j.jelekin.2014.09.008

Advances in functional electrical stimulation (FES).

This review discusses the advancements that are needed to enhance the effects of electrical stimulation for restoring or assisting movement in humans ...
2MB Sizes 3 Downloads 13 Views