69] characterized torso stiffness in male and female patients using the RoSE dynamic brace.

Several other smart and active devices have been developed to characterize different physical parameters of the spine. These devices either help improve the scoliosis treatment or enhance the physiotherapy/training potential of the patients. Khan et al. ^[70][71][70,71] have developed the cable-driven trunk support trainer (TruST), which is helpful for the training of the seated posture for patients suffering from musculoskeletal and neurological disorders, where they have compromised postural stability. Based on a patient'’s maximum trunk excursion, TruST control algorithms create a circular planar force tunnel around the trunk and provides as-needed assistance forces to the torso while performing the intended movements [72].

People with trunk impairment do not have enough strength in their trunk muscles to keep an upright posture or control the weight shifts to perform certain movements. Several passive orthoses are available to support the trunk by passively placing the torso and not allowing any degree of freedom to the trunk. Ophaswongse et al. [73] developed the wheelchair robot for active postural support (WRAPS). WRAPS supports the torso'’s weight and is capable of replicating the patient'’s trunk range of motion (tROM) without full activation of the trunk muscles. This is crucial for the people who do not have trunk control in antigravity postures, such as standing and sitting, when a reaching task is executed [74].

Table 2 describes the devices that are being used either to treat scoliosis effectively or for the training of the torso to enhance exercise potential. Several parameters, such as device name, actuation type, structure, application, and others, have been considered in the table.

3.1. Mobility and Actuation Technology

Mobility is key when it comes to treatment with braces. Conventional braces are rigid, passive, and do not allow mobility to the spine, which results in spine stiffness. Mobility can be achieved by either passive or active actuation. Actuation technologies in the area of assistive orthoses are quite matured with regard to achieving the goal of assisting the spine. Some of these actuation technologies have been introduced in corrective orthoses as well, such as the development of Spinecor and ScoliSMART, which are passive soft orthoses and use elastic material to apply corrective forces with the dynamic movement of the human body. Unfortunately, these passive braces do not provide force control over specific vertebra to apply the corrective forces.

To address the issues of mobility and apply controlled corrective forces over specific vertebra, the University of Colombia developed the RoSE dynamic brace. It uses two parallel Stewart platforms to apply forces in all directions, as shown in Figure 2. Forces and displacements applied by the RoSE can be measured by built-in pressure sensors and position sensors inside the actuator [17]. The RoSE exoskeleton is a big advancement in the area of active corrective orthoses, but has certain limitations. It uses eight series actuators that require a significant amount of power and also increase the device weight. This could be a limit in the use of RoSE, since these braces are supposed to be worn 18 h a day.

Figure 2. Smart rehabilitation orthoses for the spinal column.

Spondylolisthesis is another type of spinal deformity, in which one of the vertebrae in the spine slips out of the proper position onto the vertebra below it, putting pressure on the nerve and a disc. Atlas Japet [13] developed an active disk decompression device that particularly helps to relieve people suffering from back pain caused by a herniated disk. Excessive vertebral compression of the intervertebral disks results in back pain. Atlas Japet alleviates the pain by applying distraction force on the vertebral column to increase the inter-joint space between vertebras. Exo-dynamics developed an active spinal brace called ExMS-1 [57], which offers automatic, customizable back support without restricting mobility. The goal of ExMS-1 is to prevent back pain from becoming serious spinal injury.

Both Atlas Japet and ExMS-1 use series actuators that consume a lot of power. It is important to explore other actuation technologies that consume less power to reduce metabolic cost, as braces need to be worn for longer durations. One such mechanism is a twisted string actuation (TSA) [58]. TSA is translational transmission systems based on twisted strings coupled with electric motors, and results in lightweight, compact, and mechanically simple actuators [59]. TSAs are being used in a variety of wearable applications and have been used recently in spinal assistive devices [60]. Therefore, they have great potential to be used in the development of active dynamic braces.

3.2. Sensory Designs and Parameter Characterization

To treat a spinal deformity effectively, it is important to measure the physiological and mechanical parameters of the torso. Measuring the progression of the spine and adjusting the brace accordingly will expedite the process of recovery. Green Sun Medical developed a brace that provides physicians and patients with real-time performance metrics, utilizing a cloud-based health platform [61]. Measuring muscle activities would give feedback on muscle activation during the bracing time. This information is crucial from a physiotherapy point of view, in order to monitor the strengthening of the muscles. Myontec developed intelligent clothing technology to monitor the muscles'’ activities, integrated with movement sensors for sports and rehabilitation purposes [62].

Outcomes of the brace treatment are associated with how effectively a brace is being worn. Force sensors and compliance monitors have been developed to monitor the quality of the brace usage. Thermobrace is a temperature sensor that is applied to the brace to monitor its actual wearing. It helps to optimize the therapies and helps doctors to understand the real effectiveness of the braces ^[11][63][64][11,63,64]. One of the key concerns in brace treatment is the lack of information on the forces that are being applied by the brace on the torso. Lou et al. [65] designed a wireless sensor network system to determine the biomechanics of spinal braces and continuously monitor the forces exerted by the brace on the spine. This system helps to examine the force distribution inside the brace during daily activities.

The effectiveness of the brace treatment depends on how the brace has been worn. It is important to wear the brace with the prescribed tightness to achieve a better outcome from the treatment. Lou et.al [66] developed an active intelligent brace system, which maintains the interface pressure in a prescribed range between the body and the brace. The brace uses an air bladder, which inflates to control the pressure between brace and body. An active intelligent brace increases the effective time of wearing the brace in prescribed tightness from 28% to 47% ^[66][67][68][66,67,68].

Braces correct the abnormal posture of the spine by applying several displacements at different levels of the torso. This is usually attained by adding soft pads or by adjusting the geometry of the brace design. These displacements result in corrective forces that are transmitted to the spine through soft tissues and the rib cage within the torso. Therefore, the amount of the curve correction depends on the stiffness of the torso, i.e., the stiffness of the soft tissues and stiffness of the spine. The stiffness characteristics of the torso may vary over time and during the course of the treatment, due to variation in the torso geometry, bone maturity spine curve, and fat distribution. Therefore, it is important to consider the torso stiffness characteristics along with the spine geometry to effectively design a brace. Park et al. [17] and Murray et al. [

3.3. CAD/CAM and Smart Materials

Three-dimensional printing revolutionized the conventional way of constructing braces. For decades, the conventional way to fabricate braces was by plaster cast, which involved manual measurements of the patient'’s torso and designing a handmade brace. Technological revolution has enabled technicians to take body measurements using a photogrammetric scanning system, which is even faster than laser scanners [76]. Photogrammetric scanning systems have been proven to be effective for the fabrication of custom-made spinal orthoses, especially for patients with mobility impairments, as it allows them to capture instantaneously the torso shape with high accuracy (<1 mm) [77]. This allowed the fabrication of the CAD/CAM braces using 3D printing. These 3D-printed braces solve the socio-economic barrier of typical braces, and seem more appealing to the patients. Hence, the braces maximize patients'’ willingness to wear a brace on a daily basis. It provides unparalleled personalization, incredible breathability, and reduces fabrication time. The Flexpine brace [52] is a semi-rigid brace used for conventionally treating scoliosis. It is 3D-printed brace, which uses 4 mm-thickn foldable plastic as its frame and elastic bands to apply corrective forces. It offers great mobility to the spine and overcomes the limitations of typical rigid braces. Various other braces, such as Boston, Chêneau, and Lyon, are also being 3D printed to enhance the breathability and reduce the fabrication time.

Several studies have demonstrated significant improvement in the results using CAD/CAM braces compared to the traditional approach, adding concrete scientific evidence (level of evidence II) ^[78][79][80][78,79,80]. However, these studies cannot offer a prior guarantee for better treatment results, as several other factors play a crucial role in brace treatment. Cobetto et al. ^[81][82][81,82], in two randomized controlled trials (RCTs), concluded that the combination of finite element modelling (FEM) and a CAD/CAM approach can further improve in-brace correction (IBC). The FEM braces exhibited 48% and 47% IBC for lumbar and thoracic curves, respectively, compared to 26% and 25% of CAD/CAM braces. Axial rotation correction of 46% compared to 30% by CAD/CAM braces. Moreover, the FEM braces were 50% thinner and had 20% less covering surface, making them more breathable for the wearer ^[81][82][81,82].

Various 3D printing techniques have been adapted in medical applications, such as stereolithography (SLA), Polyjet modeling (PJM), selective laser sintering (SLS), and fused deposition modeling (FDM). High equipment costs (especially for SLS and PJM) and processing times are the major limiting factors in the production of large orthotic devices using 3D printing. FDM is the most suitable and least expensive method to produce scoliosis braces among these 3D printing techniques, despite having a bit less dimensional accuracy. A few studies have recommended the use of 3D printing to construct orthotic devices ^[83][84][85][83,84,85]. The majority of the researchers have fixated on the feasibility of developing devices with accurate geometrical dimensions and desired shapes, but quite a few studies have reported the cost and mechanical performances of these devices. Furthermore, filament datasheets often refer to the mechanical properties of the bulk material before 3D printing, while the mechanical properties of the printed parts is inadequate and primarily focused on tensile properties [86]. Moreover, it is crucial to evaluate the toughness of the material through impact tests by measuring the energy absorbed for the duration of high strain rate conditions before failure. This behavior definitely varies with different production technologies, and the materials currently used for 3D printed orthotic devices do not provide the rigidity needed to correct the spine posture [85]. For these reasons, it is not yet possible to predict how the combination of virtual modeling and additive manufacturing processes affects the mechanical properties of a 3D-printed brace.

Since one of the major drawbacks of corrective orthoses is the uncomfortable rigidity that does not allow the necessary range of motion, the introduction of the principle of variable stiffness in design seems to be quite promising. This can be obtained either using new smart materials or developing specific design solutions. Smart materials have gained substantial consideration in medical applications. In particular, shape memory alloys (SMAs) are most generally employed for their super elasticity (SE) in orthopedic treatment. Chan et al. [87] developed a flexible scoliotic brace using SMAs. Among the suitable systems for variable stiffness, the jamming-based systems are emerging with a new set of possibilities ^[88][89][88,89]. Layer-jamming mechanisms have certain advantages, such as compactness, being lightweight, high resistance force, and fast reaction time. Jamming structures also possess the shape-locking capability, which can help to reduce the metabolic cost. They can be fabricated entirely using a 3D printing technique. Choi et al. [90] proposed that these structures can be used to develop the spinal assistance robot and in various other wearable applications.