Abstract
Depending on the extent of neuromuscular deficits, interventions to improve or restore upright mobility to persons paralyzed from central nervous system (CNS) disorders can include bracing, neuroprostheses employing surface or implanted neural stimulation, motorized exoskeletons, and hybrid neuroprostheses combining neural stimulation with passive or actively controlled orthotics. Neuroprosthetic interventions can often replace, augment, or surpass the function of conventional bracing and range from simple one- or two-channel systems to address focal deficits at a single joint, such as foot drop, to surgically implanted technologies that interface directly with the peripheral motor nerves to control the actions of multiple joints and muscles. This chapter reviews the current status and future development of these options, with a particular focus on the outcomes of neuroprostheses for standing, stepping, and seated trunk control and balance after spinal cord injury (SCI), stroke, or multiple sclerosis (MS), as well as new options for exercise and recreation in the community.
Keywords
Balance, Exercise, Mobility, Multiple sclerosis, Neural stimulation, Neuroprostheses, Spinal cord injury, Standing, Stroke, Walking
Outline
Neuroprosthetic Options for Restoring or Improving Ambulation 1171
Neuroprostheses for Walking After Incomplete Paralysis 1172
Neuroprostheses for Walking After Complete Paralysis 1173
Neuroprostheses for Controlling Seated and Standing Balance 1175
Neuroprosthetic Interventions for Truncal Stability and Active Sitting 1176
Application of Continuous Stimulation 1176
Context-Dependent Stimulation 1176
Future Directions and Challenges 1176
Automatic Control of Standing Balance 1177
Future Directions and Challenges 1178
Neuroprostheses for Recreation and Exercise 1178
References 1179
Neuroprosthetic Options for Restoring or Improving Ambulation
Improving truncal stability is ranked as one of the highest priorities for functional improvement by individuals with paraplegia and tetraplegia alike ( ), and within 10 years of thoraco-lumbar-level spinal cord injuries, individuals would prefer to walk again over eliminating pain or improving other bodily functions ( ). Similarly, improving walking and upper body/trunk strength are identified as very important to more than half of individuals with paraplegia, with walking as the first or second choice in nearly 60% of the population ( ). Surgically implanted motor system neuroprostheses can address these high priority issues and provide options for seated balance and upright, bipedal mobility to individuals with paralysis resulting from central nervous system (CNS) trauma, disease or dysfunction such as spinal cord injury (SCI), stroke, or multiple sclerosis (MS).
Surgically implanted neuroprostheses for exercise, standing, and transfers consisting of multichannel implanted pulse generators and stimulating electrodes ( Fig. 97.1 ) have been well accepted by individuals with low-cervical and thoracic-level injuries ( ); deployed clinically in multicenter trials ( ), they enabled recipients to release one hand from a support device to reach objects overhead, and allowed their users to stand unassisted an average of 10 min. Such systems exhibit good reliability with robust operation and functional outcomes, and continued patterns of usage on follow-up averaging six years postimplantation ( ). Implanted neuroprostheses can also allow individuals with motor complete mid- to low-thoracic injuries to step household distances independently in the vicinity of the wheelchair ( ), and enable individuals with incomplete SCI to walk faster and farther than can be maximally achieved by volitional motion alone ( ). Triggering stepping motions with physical sensors such as accelerometers placed on the body (shoes or belt), walker, or crutch tips can be customized for a given individual and automate progress through the stimulated gait cycle ( ). Furthermore, monitoring the voluntary activities of intact or partially paralyzed muscles can make walking more intuitive by integrating the actions of the neuroprosthesis with the remaining volitional substrate, allowing natural variations in step length and walking speed ( ).
Neuroprostheses for Walking After Incomplete Paralysis
The simplest orthotic intervention for ankle dorsi/plantarflexion weakness is an ankle-foot orthosis (AFO), which provides significantly beneficial effects on balance and walking ability in individuals with poststroke hemiplegia ( ) by passively improving toe clearance during swing and passively resisting dorsiflexion to increase stability and retard tibial advancement during stance ( ). Peroneal nerve stimulation elicits active contractions of the dorsiflexor muscles and is an alternative to the AFO to correct foot drop. Many commercial devices employing surface stimulation have been introduced, most notably the Odstock Foot Drop Stimulator (Odstock Medical Ltd, Salisbury, UK), WalkAide (Innovative Neurotronics, Austin, TX), and NESS L300 (Bioness Inc., Valencia, CA). Peroneal nerve stimulation is equally effective as an AFO at significantly improving walking after gait training, with no significant differences observed between the two interventions ( ). However, long term use of stimulation may improve walking endurance and functional ambulation ( ), and significantly more users prefer peroneal stimulation over AFOs ( ). In addition, peroneal nerve stimulation appears to be superior to AFOs for avoiding obstacles normally encountered during community ambulation in users with relatively low leg muscle strength ( ). Two fully implanted stimulators for foot drop were introduced in Europe, the ActiGait (Ottobock, Vienna, Austria) and STIMuSTEP (Finetech Medical Ltd., Hertfordshire, UK) systems. These devices are designed to balance dorsiflexion and in/eversion by delivering multiple independent channels of stimulation via a single epineural cuff, through electrodes inserted subepineurally to the deep and superficial branches of the peroneal nerve. Both systems have been shown to improve walking speed and distance, are well accepted by stroke survivors ( ), and have been used in patients with MS, traumatic brain injury, and incomplete SCI ( ).
Multichannel surface stimulation can address proximal muscle weakness at the knee and hip to achieve higher functional outcomes in individuals with hemiplegia than peroneal stimulators alone ( ), and restore or improve standing and walking in individuals with complete or incomplete paraplegia. Although walking speeds and distances were limited, a multichannel surface stimulation system (Parastep, Signetics, Inc.) received FDA approval for short distance ambulation after thoracic-level SCI through appropriately timed activation of the hip and knee extensors and knee/hip flexors via a withdrawal reflex. After training 32 sessions, users could walk at least 10 m with few achieving distances as far as one mile. Benefits included increased blood flow and work capacity, reduced spasticity, and other physiological benefits ( ). Their relatively high average oxygen consumption (approaching seven metabolic equivalents) and cardiovascular stress make such systems good options for maintaining physical fitness, rather than mobility ( ). Although used by more than 400 individuals worldwide, Parastep is no longer commercially available.
Neuroprostheses can be tailored to the specific needs of individuals with incomplete SCI to address the most critical gait deficits while allowing maximal utilization of remaining volitional lower extremity function. Application of surface stimulation may be most effective in the early phases of gait rehabilitation because its flexibility in electrode placement ( ), and adjustability as voluntary and stimulated strength improve ( ). Volitional walking speed improves when stimulation is integrated into gait training (a neurotherapeutic effect), and interactive stimulation during walking provides additional gains in gait speed and distance (a neuroprosthetic effect), particularly in slow walkers ( ). Stimulating the plantarflexors, in addition to the knee and hip extensors and flexion withdrawal reflex, can also increase walking speed and stride length by providing an active push-off and improving knee flexion during swing ( ). Limitations due to habituation of the flexion withdrawal reflex ( ) and the inability to access the deep hip flexors (Iliopsoas) from the surface can be obviated with neuroprostheses employing implanted electrodes ( ). The neural structures that control all major muscles of the lower extremities and trunk can be accessed with intramuscular ( ) or peripheral nerve cuff electrodes ( ) to achieve strong, isolated, and repeatable contractions of the target muscles. Customized implanted systems demonstrate better and more consistent therapeutic and neuroprosthetic benefits than surface systems ( ) and can facilitate taking a step even in individuals with excessive extensor tone ( ). Similarly, as illustrated in Fig. 97.2 , multijoint control with implanted systems can provide significant neuroprosthetic improvement in walking ability in individuals with stroke ( ) or MS ( ).
Approximately 90% of individuals with MS will eventually experience impaired mobility ( ). Despite the prevalence of gait dysfunction and the absence of meaningful interventions to correct it, application of neural stimulation in this population is only just beginning to be explored. Small studies using surface stimulation typically limited to the peroneal nerve have demonstrated improvements in walking speed, distance, and various measures of quality of life ( ). However, other studies of similar caliber have either shown that surface stimulation of the peroneal nerve is not better than exercise alone, or confirmed a significant neuroprosthetic effect on walking performance without carryover to voluntary function ( ). The absence of a meaningful therapeutic carryover suggests that activating the proximal lower extremity musculature is likely to be critical to improving ambulation in the community and during activities of daily living for patients with MS. This can be achieved with a multichannel, implantable system comparable to what has been utilized in patients with SCI. Case studies of interventions targeting multiple joints with implanted neuroprostheses suggest that activating the hip and knee flexors (sartorius, tensor fascia latae) in addition to the peroneal nerve can increase maximal walking distance eightfold over volitional function alone ( ), and produce stepping robust and consistent enough to reduce Expanded Disability Status Scale (EDSS) scores from 7.5 to 6.0, thus preserving independent mobility and delaying wheelchair dependence.
Neuroprostheses for Walking After Complete Paralysis
Hybrid Neuromechanical Systems
Unlike walking with partial paralysis from incomplete SCI, stroke, or MS that can be improved with a relatively small number of stimulus channels to activate the muscles responsible for focal gait deficits, complete paralysis often requires more than 16 stimulus channels to control all the muscles necessary to restore stable stepping in the absence of extensive external bracing ( ). Although the maximal walking speeds and distances reported for implanted lower-extremity neuroprostheses approach those typical for community ambulation, high metabolic energy consumption (up to four times normal) and reliance on the upper extremities for balance, support, and propulsion all limit stepping with these systems to the vicinity of the wheelchair or for exercise ( ). To address these issues, hybrid neuroprostheses consisting of neural stimulation and external bracing are being developed. Early hybrid systems combined alternating activation of the rectus femoris (hip flexion) and hamstring (hip extension) muscles via surface stimulation with a reciprocal gait orthoses (RGO) that coupled the hips so that extension of one joint via volitional truncal extension resulted in flexion of the contralateral joint ( ). With knees locked and ankles fixed at neutral, this simple hybrid neuroprosthesis reduced energy expenditure 15%–30% and increased maximal walking distance from 100 to 800 m compared to the RGO alone. Exercise with the system produced positive effects on bowel/bladder function, endurance, back/muscle pain, energy, weight, spasticity, skin problems, and range of motion ( ).
In spite of significantly improved posture and trunk stability, the locked knees and reciprocating mechanism limited both step length and walking speeds (0.05–0.37 m/s), and required excessive upper body effort compared to walking with stimulation alone ( ). This motivated development of dynamic hybrid systems with context-dependent hip and knee joints that could be locked for stability in stance, unlocked for free joint movement during swing, or coupled to simulate a reciprocator at the hips ( ). Movements in these systems are powered entirely by multichannel implanted stimulation coordinated with the joint mechanisms via microprocessor-based controllers acting on the external knee ( ) and hip ( ) mechanisms to progress smoothly through the phases of gait. This significantly decreased forward lean compared to neural stimulation alone, and reduced maximum upper extremity forces on the walker by 42% and 19% compared to walking only with a standard RGO or neural stimulation, respectively. Walking speed also increased by 15% as compared to an RGO ( ).
Powered Exoskeletons
Powered exoskeletons have become commercially available to provide upright mobility for people with paralysis. Motorized knee and hip joints support and move the lower limbs, and voluntary upper body effort is exerted to maintain balance or initiate the next step by volitionally leaning forward. Notable motorized exoskeletons include Rex (Rex Bionics, New Zealand), ReWalk (ReWalk Robotics, Israel), HAL (Cyberdyne Inc., Japan), Ekso (Ekso Bionics, USA) and Indego (Parker Hannifin, USA). A systematic review of available powered exoskeletons indicates that an average speed of 0.26 m/s and maximal distance of 171 m can be expected after an average of 20 training sessions, with individual outcomes related to age, injury level, and amount of training. Mean walking speeds are less than 0.31 m/s, respectively, with the fastest speed (0.71 m/s) achieved by an individual with motor-incomplete SCI using a ReWalk ( ). These outcomes can be sufficient for supervised household ambulation in controlled environments, but generally fall short from unstructured community ambulation. While secondary benefits on pain, spasticity, and bowel/bladder function are still being determined, powered exoskeletons remain primarily exercise devices with secondary physiological benefits due to lack of active contractions of the large lower extremity muscles ( ).
Future Directions and Challenges
New approaches to walking after complete motor paralysis include combining hybrid and powered exoskeletal strategies and integrating neural stimulation with active robotic assistance. Attempts have been made to add surface stimulation to commercially available powered exoskeletons, and several developers have proposed utilizing neural stimulation as the primary source of motive power in a hybridized approach, with small motors at the joints only for assistive power as needed when stimulated contractions are insufficient to complete a maneuver ( ). Challenges to this “muscle first” strategy include seamlessly integrating stimulation-driven neuromuscular contractions with powered motor control, as well as developing more effective means of coordinating stimulation with remaining voluntary motion. With the advent of femoral nerve cuff electrodes, knee extension strength is less of a limiting factor to standing and ambulation than strong and isolated hip extension. Addressing deficiencies at the hip requires development of new multicontact electrode technologies specifically for the sciatic and gluteal nerves. In all cases, but particularly in stroke and MS applications, an additional challenge involves arresting or preventing obligatory reflexes and spasticity or excessive tone which might compromise voluntary motion or stimulation-assisted walking. Rather than exciting the peripheral nerves, the challenge is to arrest conduction immediately and reversibly as needed. This could include a combination of pharmacological agents and delivery systems (e.g., baclofen pumps), or blocking conduction of sensory or motor nerves responsible for the spasticity or reflexes. A fundamental challenge to the field remains widespread access and commercial distribution of implanted lower extremity motor neuroprostheses, which currently remain investigational in nature and of limited availability primarily through grant-funded research programs.
Neuroprostheses for Controlling Seated and Standing Balance
The motor system neuroprostheses for walking described earlier operate primarily in a feed-forward (i.e., open-loop) fashion. That is, preprogrammed temporal patterns of stimulation are selected and triggered manually or by detecting the intention to execute a specific task and progress through a cyclic motion based on input from body-mounted sensors. As such, their actions tend to be discrete in nature, and generally unresponsive to the environment except for events such as foot-floor contact, crutch loading, or acceleration of the limbs, pelvis, or assistive device. Research has focused on a deeper understanding of the roll of the trunk and pelvis in the seated operator, and advanced control systems for neuroprostheses that can automatically maintain standing or sitting balance and afford recipients the option of specifying and assuming task-specific postures best suited for a given activity.
Neuroprosthetic Interventions for Truncal Stability and Active Sitting
An important prerequisite for most activities involving upright mobility after paralysis, such as standing and stepping, is a stable torso and pelvis to maintain the head, arms, and trunk. Furthermore, virtually all seated functional activities, such as bimanual reaching, retrieving objects from the floor, or propelling a manual wheelchair, are negatively impacted by paralysis of the trunk and pelvis. The earliest studies of trunk stability concentrated mainly on the examination and characterization of the passive properties intrinsic to the inactive trunk musculature, ligaments, tendons, and other noncontractile tissues ( ). Most of these focused exclusively on the changes in trunk orientation alone, with the pelvis held fixed, and ignored the impact of rotation about the hips ( ). More sophisticated approaches have since determined the relative contributions of both the extrinsic (active muscular contractions) and intrinsic (passive tissues) mechanisms to trunk stability in able-bodied individuals ( ) as well as in individuals paralyzed by SCI ( ). Initial steps in designing advanced control systems for stabilizing trunk posture involved assessing the strategy used by the intact nervous system to mediate the static and dynamic behaviors of the trunk in able-bodied individuals ( ). Often, control system design was expedited by employing anatomically realistic musculoskeletal models of the human trunk and pelvis ( ). This allowed for extensive exploration of the feasibility of assuming various statically stable seated postures and withstanding applied perturbations by stiffening the torso with continuous contractions of the paralyzed paraspinal and gluteal muscles.
Application of Continuous Stimulation
Biomechanical simulations and experimental work with recipients of implanted neuroprostheses, including with intramuscular electrodes inserted at the spinal nerves of T12 to L2, to activate the segmentally innervated paraspinal muscles have demonstrated that continuous activation of the paralyzed hip and trunk muscles ( Fig. 97.3 ) in volunteers with SCI can improve vertebral alignment, restore a more natural anterior pelvic tilt, increase the forces that can be exerted on objects in the environment, and expand bimanual seated workspace and forward reach ( ). Postural changes with stimulation of the hip and trunk muscles can improve ventilation volumes by providing the lungs more room to expand, thus allowing the natural contractions of the diaphragm and intercostal muscles to be more effective. Stiffening the trunk by activating the lumbar erector mass can also facilitate turning in bed for individuals with low-cervical injuries by mechanically coupling the pelvis to the shoulders so that voluntary rotation of the upper body is transmitted more efficiently to the lower extremities ( ). Furthermore, stabilizing the pelvis and trunk with continuous stimulation can improve manual wheelchair propulsion efficiency and allow the shoulders to better transmit forces to the wheel rim without the need to simultaneously control a compliant spine ( ).
Context-Dependent Stimulation
Varying activation in response to internally generated or externally applied perturbations enables further functional benefits over continuous stimulation of the spine and pelvis extensors. Stimulated hip and lumbar trunk extensors are strong enough to return neuroprosthesis recipients to upright sitting positions from fully forward flexed postures, potentially obviating the need for restrictive belts, straps, or specialized seating systems. A simple threshold applied to the output of a tilt sensor on the sternum can automate the activation of this “righting reflex” ( ), or accelerometers placed on the wheelchair can detect collisions with obstacles and restabilize users to help prevent injurious falls. Research also indicates that closed-loop control of the trunk extensor muscles using a proportional, integral, and derivative (PID) and other feedback controller paradigms have the potential to greatly improve trunk stability over continuous activation by automatically modulating stimulation to maintain dynamic sitting balance in the presence of voluntary arm movements and changes in externally applied loads, even at nonerect forward or side-leaning postures ( ).
Future Directions and Challenges
Continuing work in this area includes the development, evaluation, and deployment of advanced closed-loop control systems that enable users to deploy their trunks to any arbitrary task-dependent sitting posture, automatically maintain balance at the new set-point as they reach and manipulate objects, and return to an erect posture after completing the desired activities. The clinical deployment and acceptability of such systems are likely to be dependent on the availability of implantable, rather than body-mounted, sensors and a means to standardize the insertion procedure of the stimulating paraspinal electrodes to ensure consistency and selectivity of the stimulated responses.
Automatic Control of Standing Balance
Both conventional orthoses and powered exoskeletons can allow users to stand passively with minimal upper extremity effort if the center of mass remains within the base of support, but neither is currently capable of making active adjustments to automatically maintain balance. Continuous activation of the trunk, hip, and ankle extensors can produce similar results, and allow implanted neuroprosthesis users to stand with as little as 10% body weight applied to a walker or support device for balance ( Fig. 97.4 ). A current area of motor system neuroprosthesis development involves automatically modulating stimulation to maintain a standing posture and compensate for internally generated (volitional arm movements) or externally applied disturbances. This requires monitoring such quantities as the positions, velocities, and accelerations of the joint angles, center of pressure location, or whole body and/or limb segment centers of mass and regulating them via intelligent control systems.
Related posts:
Fundamentals of Burst Stimulation of the Spinal Cord and Brain
Advances in Invasive Brain–Computer Interface Technology and Decoding Methods for Restoring Movement and Future Applications
Neurostimulation for the Treatment of Complex Regional Pain Syndrome
Deep Brain Stimulation of the Pedunculopontine Tegmental Nucleus Improves Static Balance in Parkinson’s Disease
Colonic Electrical Stimulation for Constipation
Transcranial Alternating Current Stimulation and Transcranial Random Noise Stimulation
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
