Perspective Chapter: Hardware Technologies for Gait Restoration

Abstract

This chapter summarizes current status and perspectives in hardware technologies for medical rehabilitation. The chapter’s first part describes the long journey from basic rehabilitation technologies toward modern robotic devices for gait restoration. The second part of the chapter revolves around a context-based use of hardware techniques: their connection with rehabilitation goals, International Classification of Functioning (ICF) coding, rationale for use, and combinations with other treatment modalities. This part summarizes the opportunities of modern hardware gait and highlights their limitations, both clinical and organizational. The third part revolves around available evidence-based data on the comparative effectiveness of rehabilitation technologies in different clinical scenarios. The final subchapter describes the state-of-the-art hardware restoration techniques, including implanted electrodes, exoskeletons etc., as well as gives an outlook toward the challenges faced with researchers and healthcare professionals.

1. Introduction

Rehabilitation of gait stereotype is one of the most well-established medical rehabilitation interventions with high chance of clinical success. This is due to the large number of rehabilitation technologies and training complexes that have been successfully introduced into clinical practice over the past 25 years [1, 2, 3, 4].

Before the first robotic hardware devices for locomotion therapy were introduced in 2001, the restoration of the walking stereotype was carried out by physical rehabilitation professionals using “manual” methods. This was an extremely labor-intensive task and required not only skill but a great deal of stamina. Manual rehabilitation of the patient’s walking was usually carried out on bars or on a “treadmill”, with the involvement of at least two specialists who provided “manual” assistance in moving the patient’s legs. Of course, this technique also had a number of requirements on the patient’s side. These requirements included moderate degree of movement disorders, positive postural control, and cognitive and compliance levels sufficient to follow the recommendations during training. The basic means of rehabilitation were four-leg walkers (with high and low support levels), rollators, canes, crutches, single-support canes, the use of which was taught to patients during the course of rehabilitation [5, 6].

This combination of “manual” gait training and the use of technical means of rehabilitation for external support during walking did not allow to effectively restore the physiological stereotype of gait, but rather contributed to general locomotive function.

Modern arsenal of rehabilitation devices and technical means of rehabilitation is extensive. In principle, all technical devices used in practical rehabilitation can be divided into four large groups:

1. systems for unloading body weight;

2. biofeedback systems—reconstruction;

3. robotic walking recovery systems;

4. medical load suits.

Effectiveness and utilization of these technologies per se or in combination with other rehabilitation modalities like electric or magnetic stimulation vary greatly between different clinical contexts, treatment strategies, and rehabilitation prognostic groups; moreover, different aspects of gait can be improved with the use of these techniques [7]. It must be stressed out, that these technologies are not mutually exclusive and can sometimes be combined to treat a single patient. Moreover, sometimes they are intentionally coupled to use their specific traits in combination (e.g., robotic biofeedback systems) [4].

2. Overview of hardware rehabilitation technologies

Body weight unloading systems are devices that allow rehabilitation professionals to securely fix the patient in upright position using a suspension system. Such systems usually include a vest with clamps in the groin area that are tethered to vertical support posts located either above the patient’s head or in the side support posts [8, 9, 10].

All body weight unloading systems are divided into stationary and mobile models. This division addresses mostly the devise’s intended modality of use, not it is technical specification.

The stationary models of the body weight unloading system is a complex technical solution that combines the body weight unloading system and the rehabilitation “treadmill” into a common simulator, with a minimum belt speed of 0.5 km/h, an emergency stop system, often with the ability to read heart rate. These devices are usually installed in medical rehabilitation units. Examples of such devices are “Ortorent C +”, “Smart Gravity”, “Body weight support system”, and “others.”

Mobile body weight unloading systems represent a mobile parapodium, with patient fixation and the ability to exercise the patient’s walking training directly in the ward or the corridor of the department. As a rule, the appointment of these devices is consistent and at the beginning of the rehabilitation course, the patient begins walking training in the early period in the department, after expanding the motor regimen, the patient is transferred to further training in rehabilitation halls on a stationary simulator with unloading of body weight. These devices are used to tackle two challenges of early rehabilitation—increased risk of falling and need for dosed axial unloading from the lower extremities during walking training, which is often required, especially in patients after musculoskeletal surgery procedures or asthenic patients. Examples include “Ortorent С”, “Ortorent M”, “Andago”, “Aceso”, “EGO”.

A different ideology of the same technological decision is provided by anti-gravity rehabilitation treadmill “Alter-G”, which allows a controlled reduction in the load on the musculoskeletal system, ranging from 20% to 80% of the patient's body weight thanks to differential air pressure technology. During the procedure, the patient enters a special chamber, after which the specialist sets the speed of the walkway, the slope, and air pressure, which gradually increases, gradually removing the axial load on the patient’s lower limbs in 1% increments.

In addition, ceiling-mounted “rail” systems for fixing and unloading body weight should also be included in this category of technologies. Often, both patient rooms and rehabilitation rooms are equipped with ceiling “rail” systems. The “rail” system includes a mobile free-sliding “node” to which a patient support device, typically a vest, is attached. Modern rail systems used in the rehabilitation of walking patients are provided with “knots” with a shock-absorbing function that allows physiological training and does not deprive a securely fixed patient of the possibility of physiological training of step elements in conditions of sufficiently rigid vertical fixation. Examples of a “rail” ceiling patient fixation system include GTS (Mega Sarana Medica), Float (Reha Stim), and PRM-01.

BFB-reconstruction systems, in contrast with body weight unloading systems, not only allow for training of the patient’s walk skills while maintaining safe fixation and partial body weight unloading but also restore the physiological stereotype of walking through training the elements of the step and walking characteristics in an active motor mode [4]. The most striking example is the C-Mill rehabilitation complex for biofeedback video reconstruction of walking. This rehabilitation complex allows you to play the video sequence on the canvas of the track, individualizing the training and taking into account the pathological signs of the patient’s walking. Video marks allow you to adjust such characteristics of the pattern as the length and width of the step, to train complexly coordinated walking elements. Video-audio training support allows physicians to set a goal for rhythmic steps with the same length and width of the step of the contralateral limbs. In addition, this rehabilitation system by means of an integrated podometric diagnostics system, allows the specialist to carry out not an empirical selection of a technique for training the walking stereotype but to conduct initial and dynamic testing of the patient to determine the key gait abnormality features. This examination can be followed by individualization of the walking recovery technique based on the results of the diagnostic study. The rehabilitation complex is also provided with a system for fixing and unloading the patient and means of emergency stops.

Both body weight unloading and BFB-reconstruction systems, however, rely greatly on patient’s cognitive skills and his/her ability to cooperate with rehabilitation team. They also require some initial patient’s training, and creating user-friendly biofeedback mechanisms is considered a top research priority [2].

Robotic systems for restoring the stereotype of walking. The era of automated imposition of a physiological walking pattern by mechanotherapy in robotic rehabilitation complexes was laid by the Swiss company Hocoma, a medical engineering company that develops innovative equipment for rehabilitation, the first robotic product of which was the Lokomat System [3]. This fundamentally new rehabilitation device appeared on the market in 2001 and opened the era of robotic rehabilitation devices, which has become not only a new direction in the development of high-tech rehabilitation care but also made it possible to train walking in patients with profound motor disorders, including patients with lower paraplegia, hemiplegia, and tetraplegia, which was impossible before the advent of this equipment. Thus, the creation of the Lokomat System, after many years of research and development, has brought a significant breakthrough in the field of locomotor therapy. The Lokomat System is a hardware complex consisting of three parts: external robotic exoskeletons, structurally imitating the lower limbs (motorized nodes reproduce the hip and knee joints, and the “stirrup” fixes the patient’s foot). The technical solution of the exoskeleton allows not only to provide rigid fixation of the legs of patients but also to set the necessary anthropometric, goniometric, and speed characteristics, thereby individualizing the training procedures, taking into account the anatomical and physiological characteristics of each patient. The patient unloading system provides maximum technical safety for patient verticalization, as well as the possibility of providing dosed axial unloading. The constructive solution of the treadmill path, along which the patient walks, allows you to start training at a minimum speed (1 km/h). The operation of the exoskeleton is synchronized with the speed of the track. The software allows not only walking training in the passive mode but also dosed to reduce the degree of functional activity of the robot (alternately or simultaneously) from the patient’s legs, thereby conducting training in the active motor mode, subject to the imposition of a physiological walking stereotype on the patient. Currently, in clinical practice around the world, there is a sufficient number of analog robotic systems: robotic complexes “GEO” and “A3”. The principle of operation of all robotic rehabilitation devices is based on the imposition of a physiological stereotype of walking by exoortheses in the patient’s fixation system. However, nevertheless, there are a number of differences, so the GEO complex, unlike the Lokomat System, allows you to train walking not only in a straight line but also to simulate going up and down stairs, thereby expanding the possibilities of using the technology in patients of different nosological forms and patterns of gait disturbance.

The next type of robotic technology that has also been developed in the Hocoma industrial laboratory is the Erigo robotic stander. The turntable with an integrated stepping device has been designed to facilitate and facilitate the early mobilization of neurological patients. This hardware complex will allow for staged verticalization of patients with a given verticalization angle, and an integrated robotic stepping device that starts mobilization of the patient’s lower extremities in a horizontal position allows for more effective verticalization with a reduced risk of collaptoid conditions. The device securely fixes the patient in a corset in the body weight unloading system on the horizontal surface of the verticalizing table with setting the verticalization angle; the integrated robotic stepping device fixes the distal parts of the lower extremities and allows the specialist to set the walking speed and rhythm parameters. The software creates and saves individual reports of verticalization procedures, allowing for dynamic monitoring of the effectiveness of the ongoing course of procedures with a logged report form. An analog example of this type of equipment is the hardware complex “A1”.

Exoskeleton (from the Greek. ε’ξω—external and σκελετος—skeleton) is a device designed to increase human strength by means of external frame. Initially developed for the military, they now reliably entered the rehabilitation and habilitation practice. There are models of exoskeletons with an active and passive principle of operation—active and passive exoskeletons. Active models use external devices as a source of energy, while the mechanics of passive exoskeletons are based on the use of kinetic energy and human strength. Active exoskeletons are widely used for military purposes, but their performance data is rarely open to public or non-military researchers. The maximum number of such developments falls on the Pentagon. One of the well-known exoskeletons HULC (Lockheed Martin, USA), allows a soldier to quickly move with a load over rough terrain, while there is a high speed of movement. HULC helps not only carry but also lift the load from the ground. The mass of the device is 25 kg, most of it falls on batteries, and the battery lasts for two hours. From a medical perspective, their number also increases; medical exoskeletons can be divided into two groups.

The first group completely the ReWalk exoskeleton (ARGO Medical Technologies, Israel) allows patients with lower paraparesis to stand up and walk using sticks. The operation of the structure is based on sensors that detect the forward tilt of the body and transmit a signal to the devices supporting the legs. Power is supplied from a battery placed in a special backpack behind your back. The design can only be used in persons with preserved functions of the upper limbs. REX (REX Bionics, New Zealand)—provides additional support for the human body when moving. Management is carried out using a joystick and a tablet. The mass of the exoskeleton is 38 kg, which, together with the high cost, greatly limits its wide-scale use. HAL—Hybrid Assistive Limb (Cyberdyne, Japan)—is designed for the elderly and disabled people who have difficulty in moving. The total weight of the structure is 23 kg, height is 160 cm. In addition, the battery weighs 10 kg, and the battery life (under maximum load conditions) is 2.5 hours. These products mostly focus on patients that have little to no perspective of returning to normal gait stereotype either due to excessive skeletal trauma or as a result of spinal cord disruption or specific brain trauma.

eLEGS (Ekso Bionics, USA) is a special hydraulic exoskeleton designed for patients with lower paraparesis. The design allows them to move around with the help of crutches or special walkers. The heart of the machine is an interface-hardware-software complex that uses natural human movement to safely translate it into exoskeleton action using a microcomputer. Passive exoskeletons have found their primary use in military applications. In Russia, the Transport Walking Systems company has created a passive exoskeleton K-2, designed for the needs of the Ministry of Emergency Situations. This device helps a person to carry loads weighing up to 50 kg for a long time without much effort and load on their own musculoskeletal system. A group of Russian scientists from the Research Institute of Mechanics, Moscow State University. M.V. Lomonosov, a working sample of the ExoAtlet P passive modification exoskeleton for rescuers of the Ministry of Emergency Situations has been created, which allows the human operator to carry large loads (70–100 kg). The use of exoskeletons in practical rehabilitation is quite widespread; however, it is still limited due to the severity of the device and the difficulty of walking in it, taking into account the need to maintain balance while relying on crutches. Thus, exoskeletons should be considered to a greater extent in the aspect of assistive technologies for spinal patients as an opportunity for their mobilization and habilitation. At the same time, stationary robotic complexes such as Lokomat are more promising for training a dynamic walking pattern. At the same time, it should be noted that in pediatric practice, on the contrary, the use of the ExoAtlet Bambini exoskeleton is seen as extremely promising for the rehabilitation of children and adolescents with locomotion disorders. Training in a given physiological pattern of walking with the help of a specialist, without additional support on crutches, in combination with the lightweight design of the exoskeleton itself, makes it possible to conduct locomotor training of children in an automated imposition of a stereotype.

The systems of robotic imposition of the walking stereotype also include simulators with ellipsoid step automation. These robotic step simulation systems differ from the previously described complexes by their smaller overall dimensions, lower cost, and ease of operation. The simulators represent an automated simulator of the “ellipse” type, with a system of dynamic unloading of body weight and software that allows you to implement the stereotypical act of automating the patient’s walking forward and backward, at a given walking speed and taking into account the patient’s anthropometric parameters. Training complexes have a built-in control of biological indicators, control of the angle of elevation of the step, step length, as well as resistance during movement. An example of a training complex: “Ortorent С++”.

As clinical, academic, or pragmatic guidelines do not exist for use of exoskeletons, there is much discussion on their role and effectiveness in clinical practice. Recent studies suggest exoskeletons can benefit inpatients with gait impairment [11].

The fourth type of rehabilitation equipment is therapeutic power suits, which is a powered system consisting of supporting elements and adjustable rods, with the help of which, for a therapeutic purpose, a load is imposed on or lifted from the musculoskeletal system [12]. This type of rehabilitation device was created on the basis of developments in space medicine and, in particular, the Penguin suit, designed to protect astronauts from the adverse effects of weightlessness. Therapeutic power suits are both a soft orthopedic apparatus and a load simulator. As an orthopedic device, it contributes to the simultaneous correction of the posture, bringing the joints to the maximum physiological position. As a load suit, it helps to extinguish pathological reflexes and dose the load, enhancing the effect of therapeutic exercises. An extremely important therapeutic effect from the use of costumes is the elimination of the pathological tone that prevents the implementation of the physiological locomotor act. These medical suits are widely used in neurorehabilitation, especially in pediatrics. An example of this technology is the Adele medical suit. Another example of a medical suit is the reflex-loading device “Gravistat”. Numerous rods set an adjustable compression load directed along the long axis of the body and correct the position of individual motor segments of the trunk and lower extremities. Overcoming additional resistance increases the activity of postural muscle groups. Axial load and functional correction of the position of motor segments of the body by rotational rods leads to the emergence of impulses from the receptors of the articular-muscular-ligamentous apparatus to the central nervous system. The third representative of the family of medical suits is the neuro-orthopedic rehabilitation pneumosuit “Atlant”, developed on the basis of the high-altitude compensating flight suit VKK-6, which was part of the equipment of pilots and astronauts. RPK “Atlant” promotes polysegmental stretching of the muscular-ligamentous apparatus and activation of the motor-neuronal system at all levels of the central nervous system. The Atlant suit creates neurophysiological conditions that allow the patient to maintain a posture, perform voluntary and coordinated movements.

3. Clinical use context

An important aspect of the rehabilitation of walking is the definition of its place in the structure of the International Classification of Functioning (ICF), for the correct positioning of the role of mechanotherapeutic restoration of the locomotor act in complex programs for the rehabilitation of patients. Gait disorders are specified as the third detailing level in ICF and are defined by four main domains:

• d4500: walking short distances. According to the WHO comments on domains—walking for distances less than a kilometer, for example, in a room, corridor, short distances outside the home;

• d4501: long distance walking. According to the WHO comments on domains—walking distances of more than a kilometer, for example, from one part of a village or city to another;

• d4502: walking on various surfaces. According to WHO comments on domains—walking on sloping, uneven, moving surfaces such as grass, ice, gravel, snow, walking on the deck of a ship, train, or other vehicle;

• d4503: walking around obstacles. According to WHO comments on domains—walking between moving and stationary objects, among people, animals and vehicles, walking in the market shop, in traffic conditions or in crowded places.

An important element of gait rehabilitation in order to individualize training methods is the diagnosis and analysis of the pattern, both at the beginning of the course and in the dynamics of training [13]. For example, the use of the diagnostic tool of the previously described C-Mill biofeedback videoreconstruction complex allows specialists to carry out the necessary diagnostics without additional equipment and with relaxed skill requirements. However, pre-installed diagnostic tools are found only in selected hardware complexes, so it is usually necessary to take into account the need for diagnostics using the following technical means and conditions:

• a sufficiently large free area to assess the free gait of the patient (at least 6 meters in length and 3 meters in width);

• “running” track;

• stopwatch for recording training time and fixing heart rate;

• a video camera and a computer for “cutting” video fragments;

• software for 2D analysis;

• plantar pressure recording system;

• power platform and vector video display unit;

• four- or eight-channel surface myographic amplifier;

• 3D motion analysis system;

• breath analyzer.

Rehabilitation goals for walking restoration programs should be determined by the domains of the ICF. Increasing the distance and speed of walking, as well as, and these are priority goals for restoring the functioning of the patient and improving his quality of life—this is the development of complexly coordinated elements of walking, and overcoming obstacles and confident walking on different surfaces, including unstable ones. All of the above tasks should be realized under the condition of restoring the elements of the step and restoring the global physiological stereotype of walking. These goals, when compiling complex rehabilitation programs, are formed sequentially: first, the distance and speed of walking are trained, then the tasks of adapting the patient to the social and domestic environment and training walking with obstacles on various surfaces, in combination with the implementation of complex functioning skills, are already solved, including the lack of visual control over the implementation of walking. In this regard, an important component in the staged implementation of walking recovery is the inclusion of ergotherapeutic walking training technologies in the rehabilitation program, namely the restoration of walking, as a component of the implementation of a certain social, household, or labor skill. An example of such a complex rehabilitation with recovery according to the domains of the ICF on the functions of overcoming obstacles and walking on various surfaces was created by the author’s team. In this training device, the BFB principle is implemented by distracting the patient’s focus on steps with audio and visual. In this rehabilitation complex, the multidimensional reproduction of social and domestic tasks is carried out by modeling various situations either on the screen in front of the patient or using VR technologies with the patient’s maximum immersion in the “real” environment. At the same time, both ergotherapy items (locks, door handles, magnetic card readers, etc.) and unstable surfaces that make it difficult to walk on “reference” rehabilitation surfaces are implemented in the complex training system [14].

When discussing the means of walking rehabilitation, one cannot fail to note the importance of a staged and integrated approach to the issue of restoring this complex locomotor act. The phasing of the use of mechanotherapeutic means of rehabilitation is reflected in the successive change of rehabilitation devices, taking into account the degree of verticalization and the possibility of reproducing motor acts in an active motor mode. As a rule, the following sequential system of using the described equipment occurs. Hardware verticalization, followed by the appointment of either robotic systems (in case of deep motor deficits) or biofeedback video construction in active motor mode (in case of a shallow motor deficit), with a gradual transition to body weight unloading systems for walking training without the condition of imposing a physiological pattern. This staged approach is more typical for patients with neurological deficits, while patients with lesions of the musculoskeletal system, as a rule, are limited to the appointment of devices with unloading of body weight and/or biofeedback reconstruction and undergo rehabilitation of walking in the active motor mode, subject to axial unloading of the lower extremities.

In clinical practice, the use isolated methods of mechanotherapy are not feasible. The restoration of the walking stereotype has a systematic and integrated approach [7, 15, 16]. In complex walking rehabilitation programs, in addition to individual kinesiotherapy sessions to work out individual elements of the step and strengthen certain muscle groups, balance training using stabiloplatforms, including BFB for training balance and locomotor symmetry, is prescribed, for example, “Huber”, MBN “Stabilo”, and others. A simulator that simulates climbing stairs with an electric drive is widely prescribed for training getting up and standing, for training walking up and down stairs, and for the ability to train walking on steps of different heights. An example of this equipment can be: “DST 8000 Triple Pro”, “Alterstep”. In the vast majority of cases, mechanotherapeutic rehabilitation of the walking stereotype is carried out in combination with electromyostimulation technologies. Special attention should be paid to “Walk Aid”, a programmable wearable electrical stimulator, widely used in foot paresis, which gives an impulse to the “transfer” phase in the step sequence, thereby allowing patients with peripheral paresis to carry out the physiological walking pattern. The technologies of multichannel functional electrical stimulation are also widely applicable, which make it possible to enhance rhythmic muscle activation during walking training. Examples of devices—“Akord”, “Trust-M”, “MNS-16”, and others.

4. Current evidence on best training system utilization

Many fucking devices—lots of techniques to choose. Aggregated data suggests that both the functional deficit’s properties and the clinical diagnosis contribute to the desired properties of hardware rehabilitation technique. Studies vary greatly in quality of evidence and patient groups. Thus, to summarize briefly the overall benefits of hardware rehabilitation devices, we have extracted data from meta-analysis of related studies and specifically discussed the issues of study design quality, study data applicability, and limitations. The systematic reviews are listed in Table 1.

№	Author and date	Patient group	Compared modalities	Conclusion
1	Moucheboeuf 2020 [3]	Post-stroke patients	Robotic vs. physiotherapy vs. physiotherapy combined with body weight support training	Robot-assisted training resulted in greater improvement of gait speed and balance
2	Yang, 2022 [17]	Spinal cord injury (N = 497)	body weight-supported overground training vs. body weight-supported treadmill training vs. robot-assisted gait training vs. conventional training	Each hardware modality is better than conventional physiotherapy, but no studies compare them with each other
3	Glanz, 1995 [18]	Poststroke (chronic) patients with hemiplegia	Biofeedback plus physiotherapy vs. physiotherapy alone	No increase in range of motion of speed
4	Hsu, 2023 [19]	Post-stroke, any time after the event (N = 492)	Exoskeleton-assisted training vs. conventional gait training	Exoskeleton-assisted training is superior in walking speed and balance
5	Mehrholz, 2017 [20]	Post-stroke, any time after the event (N = 1472)	Electromechanical- and robotic-assisted gait training versus physiotherapy	Hardware-assisted training was associated with increased chances to return to independent walk
6	Patathong, 2023 [1]	Incomplete spinal cord injury (N = 709)	Physical therapy vs. treadmill vs. functional electrical stimulation vs. robotic-assisted gait training,	Functional electrical stimulation is the best in regaining speed and velocity; no large studies available

Table 1.

Systematic reviews and meta-analysis on hardware rehabilitation technologies.

It is important to stress the key limitations of these studies. They can be divided into three overlapping groups: related to the nature of the underlying medical condition, related to the structure of patient management, clinical logistics, and medical aid timing, and related to the technical characteristics of the devices in question and their utilization context.

In the first group, the associated results can significantly vary with time, which is especially important in cases with specific disabling events like stroke or trauma; majority of studies describe late subacute (3–6 months) or chronic (>6 months) period after stroke or trauma; studies focusing on acute and early subacute patients are rare [2]. For example, of all biofeedback studies, only Druzbicki et al. used step-length biofeedback during bodyweight-supported treadmill walking [21].

Second, the studies’ results may be associated with specific patterns of available equipment in different hospitals. The most impressive example is the meta-analysis by Yang et al. [17], where 15 studies compared some hardware rehabilitation technology with physiotherapy, but no comparative studies between hardware modalities were found. It is also worth mentioning that specific hardware modalities themselves are quite multifarious by regimen, intensity of use and specific operating modes, making comparison between this embedded treatment options not less importan than comparison between different modalities.

Last, but not the least, technical properties of hardware rehabilitation devices and selected modalities also affect the treatment outcomes. Rehabilitation programs include not only the hardware devices themselves but also the associated gait restoration techniques, and these variations in associated training can contribute to the hardware rehabilitation results.

In the end, available comparative data on hardware gait rehabilitation techniques, especially in high-quality RCTs, is very limited. The enormous variability of devices and their use options, multiplied by the fact that hardware rehabilitation cannot be the single rehabilitation intervention in these patients, as well as the heterogeneity of patient groups, greatly restrict the quality of evidence in this field.

Studies may suggest that careful selection of patients with poststroke gait disorders for specific modalities of exoskeleton and robotic rehabilitation can improve walking speed and balance outcomes compared to modalities. In patients with spinal cord injury bodyweight-supported overground training, bodyweight-supported treadmill training and robot-assisted gait training are all more effective than conventional training, but no comparative research between them exists [17].

More specific assumptions arise not from meta-analysis but from individual studies. A large article by Mikolajczyk et al. recaps the available data on specific gait rehabilitation modalities in stroke, spinal cord injury, Parkinson’s disease, and multiple sclerosis with respect to speed, stability, and independence of walk aspects [22].

These details suggest that the following use context. In any individual patient, the gait disorder pattern should be decomposed to components specific to known benefits of available hardware rehabilitation devices (e.g., walking speed, balance, or independent walking ability). This can help the clinicians to prioritize one technology over the other for this particular patient while staying inside an evidence-based context.

In our opinion, more real-world data would suggest a more pragmatic and, in the meantime, broad look at hardware gait rehabilitation. Internet of things and artificial intelligence analysis will finally reduce the workload demands for new research, providing rehabilitation society with new quality evidence on this topic.

5. Technological perspectives

First in line of emerging technologies come exoskeletons, who is clinical role, as described above, is still a subject to discussion with a gradual increase of general clinician acceptance [11]. Their current use is limited by high costs and high weight to inpatient settings only. Upscaling of production, use of new materials, and increased battery useful life can also broaden their practical spectrum.

Minituarization offers another benefit for gait restoration like outpatient-based gait analysis devices. While they are focused mainly on measuring Parkinson’s disease gait response to treatment, its possible applications in rehabilitation, especially home-based rehabilitation (see below), could not be underestimated [7]. In the meantime, biofeedback devices earn increasing popularity for poststroke gait rehabilitation as systematic reviews give more positive results on their effectiveness [2].

A short-term technological perspective of hardware rehabilitation encompasses a broad spectrum of AR and VR technologies. As a part of the size-reducing initiative, augmented reality technologies can greatly enhance both inpatient and outpatient performance of gait rehabilitation [23]. However, literature reviews often focus on academic, rather than technological perspectives of these technologies [24].

Deep-tech perspectives of gait restoration lie mostly in the field of neural stimulation and neural implanting. Generally, both spine and peripheral installments show promising clinical results, and some randomized controlled trials are already available – for example, on peroneal nerve stimulation [25]. However, more interesting results are shown in studies, yet single patients describe brain-spine implantable devices [26]. Astonishing case reports with one complete walk restoration after complete spinal injury indicate that this stimulation devices can possibly lead to real neural regeneration – at least turning the device after a prolonged period of use did not compromise the restored functions [27].

However, even wider use of such technologies does not reduce the need for more “classical” hardware rehabilitation technologies. First, they are currently limited to use in patients with relatively low spinal injuries and preserved cognitive function. Second, the rehabilitation process still require a “training” phase, where the patient learns to control his “newly-acquired” locomotion skills. Last, but not the least, patients with leg trauma constitute a substantial proportion of all patients requiring gait restoration treatment.

Conflict of interest

The authors declare no conflict of interest.