Early onset scoliosis (EOS) is a type of spine deformity that presents before 10 years of age. The biomechanical properties in scoliosis have been found to be different, especially in the case of the concave and convex paravertebral muscles. Based on this fact, a novel three-dimensional (3D) printed patient-specific asymmetric stiffness brace design method is proposed in this paper, aiming to provide asymmetric stiffness to match “imbalanced” biomechanical properties of the concave and convex paravertebral muscles, respectively, and treat EOS by applying the block-structure brace. A 3D computer aided design draft model of the brace contour was implemented from 3D scanning. The asymmetric stiffness block-structure brace was designed in Rhinoceros and the finite element (FE) model was imported into abaqus. FE simulation was employed to study the mechanical characteristics of the brace, which provided a quantitative index for the imbalanced property of brace stiffness. The results of the FE simulation showed that the stiffnesses of the concave and convex sides were 145.88 N/mm and 35.95 N/mm, respectively. The block-structure brace was fabricated using 3D printing. Asymmetric stiffness was evaluated by corrective force measurements, which were obtained from a thin-film pressure sensor equipped on the brace. The patient-specific asymmetric stiffness brace was applied to clinical practice in a one-year-old EOS patient. A novel low-cost 3D printed brace design method for EOS was proposed in this study that could potentially be useful in patient treatment acceptance.
Early onset scoliosis (EOS), which appears before the age of 10, can be due to congenital vertebral anomalies, neuromuscular diseases, scoliosis-associated syndromes, or idiopathic causes . EOS, in contrast to adolescent scoliosis, can have serious consequences for lung development and significantly reduce life expectancy. The purpose of treatment is to manage the deformity, enable spinal and chest wall growth, lung development, and enhance pulmonary function . Nonsurgical treatment is the first course of action. If this is unsuccessful, fusionless surgery can be performed to delay the final fusion procedure until the patient is older. Nonsurgical treatments include bracing and serial casting, which are important surgical delaying tactics for a disorder where all surgical treatments have common complications. Excellent results have been reported in very young patients with curves less than 60°, with serial casting or bracing to maintain correction [2,3]. Bracing is the most commonly used treatment for scoliosis . The therapy strategy is based on the overcorrection of curves by pushing the spine in the opposite direction of the problematic curves, which is expected to result in the cessation of scoliotic curve progress. However, some of these braces are manufactured from an “off-the-shelf” module selected using measurements from the patient. Pads are added at areas needed to produce correction and trim lines are added to produce relief areas. However, the issue of compliance seen with brace wear still exists . Although a significant part of orthopedic braces is designed nowadays in computer aided design (CAD) from three-dimensional (3D) scanning of the torso shape of the patient, the asymmetric complementary balance of brace stiffness is seldom considered.
To the knowledge of the authors, no published study exists that explores brace design considering the asymmetric stiffness properties of the concave and convex paravertebral muscles in EOS. Abnormal paravertebral muscle development was thought to induce scoliosis [6–10]. Scholars have used various methods to study the relationship between stiffness of paravertebral muscles and scoliosis. Wajchenberg concluded that paraspinal muscles in the concave side of the scoliosis apex had significantly more fibrosis and fatty involution, while both sides showed signs of myopathy, muscular atrophy due to necrosis, presence of hyaline fibers and mitochondrial proliferation . The paravertebral muscles in moderate scoliosis have been shown to demonstrate a lower type IIB/IIA fiber ratio on the convex side, along with an increased proportion of type I fibers . Scoliosis is caused by unbalanced stiffness of the intertransverse ligament and the ligamentum flavum, according to a numerical study of the spine . The proportion of Type I fibers in paravertebral muscles has also been shown to be lower on the concave side, which is a finding consistent with reduced muscle activity observed in scoliosis [14,15]. Liu further illustrated the biomechanical properties of the convex and concave paravertebral muscles in scoliosis. Findings included that concave paravertebral muscle tone and stiffness were greater than those on the convex side and the asymmetric biomechanical characteristics of paravertebral muscles were closely related to the severity of scoliosis . Patients with EOS are still in a critical period of growth and development. Therefore, controlling the development of scoliosis while maintaining the growth of spine and thoracic cavity as far as possible has become the focus of therapy for scoliosis, and is also the most controversial and challenging problem in the field of spinal deformity. In this paper, a one-year-old patient suffered from scoliosis deformity combined with tethered cord syndrome and a cyst in the sacral canal. Considering the young age, surgical treatment may not be as effective as conservative orthopedic bracing treatment. Therefore, with the consent of the parents, brace treatment was used to preserve or correct the global balance of the trunk, while the patient is awaiting surgery.
This paper proposes the 3D printed patient-specific asymmetric stiffness brace design method, a novel application area division method of the corrective force and configuration for the application of the corrective force to realize asymmetric stiffness. A block structure was developed in this paper to enable the asymmetric stiffness of the brace to match the imbalanced biomechanical properties of the convex and concave paravertebral muscles, respectively, and to provide corrective force in scoliosis, while promoting therapy and rehabilitation of patients. Finite element (FE) simulation was employed to study the mechanical characteristics of the brace and act as a quantitative index for the imbalanced property of brace stiffness. The brace was equipped with a thin-film pressure sensor. Finally, the brace was manufactured using 3D printing technology and clinically applied.
2 Materials and Methods
The parents of the patient were informed that data concerning the patient would be submitted for publication, and consent was provided.
2.1 Patient-Specific Asymmetric Stiffness Block-Structure Brace Design and Fabrication
2.1.1 Data Acquisition and Correct Force Applying Area Division.
A 3D geometrical model was obtained by 3D scanning the scoliosis region from patient Xu (Female, 1Y old, 12 kg) of Xi'an Children's Hospital, China. The patient suffered from scoliosis deformity combined with tethered cord syndrome and a cyst in the sacral canal. The spine exhibited manifestations of vertebral dysplasia, including small vertebral body, osteoporosis, and occult spinal dysrhaphism, for which the surgical treatment is usually associated with multiple complications and the pedicle screw placement is often difficult in orthopedic surgery. While the child is in the first peak period of growth and development, orthopedic surgery would restrict the spinal growth and development in some degree, resulting in high failure rate for surgery. Systematic analysis of the X-ray images (Fig. 1(a), supine standing position of the patient) was performed to optimize the corrective forces (push/shift). 3D scanning (Fig. 1(b)) data were imported into a computer for the 3D model reconstruction (Fig. 1(c)). A 3D CAD draft model of the brace contour was constructed (Fig. 1(d)).
The division of the corrective force application area and the amount of brace is often judged by the experience of the brace technicians, making the formation of a unified quantitative standard difficult [17,18]. However, this process is critical in brace biomechanics correction. The effectiveness of a novel technique was hereby investigated: Specifically, Cobb angles are measured from the top of the upper-end vertebra to the bottom of the lower-end vertebra in the coronal plane in posteroanterior view . First, the line parallel from the top of the upper-end vertebra is extended and the line parallel from the bottom of the lower-end vertebra intersects with the convex side of the torso contour at two points, A, B, in the coronal plane. Second, a line parallel to the horizontal plane is drawn through A and B that intersects with the torso contour at the two points, C, D, respectively (Fig. 2). The blue area (boundary S1) and deep pink area (boundary S2, S3) in Fig. 2, where these two lines intersect with the anterior and posterior axillary lines of the convex and concave side, mark the corrective force application area (boundary S1 for convex area and S2, S3 for concave area).
2.1.2 Asymmetric Stiffness Block-Structure Brace Design.
The asymmetric stiffness brace in this study was a three-point force correction system, consisting of one action force (Fig. 3-1) and two reaction forces (Figs. 3-2 and 3-3), which provide a convex-to-concave tissue transfer.
The desired stiffness of the concave and convex sides can be achieved by the different thickness of the brace where the concave and convex side block structure is located. The distinction lies in the fact that the thickness of the concave and convex corrective force application area is different, while the thickness of the brace in the remaining areas remains the same (Fig. 4). Block structures parallel to the torso of the geometrical epidermis of the patient toward the torso direction were designed on the corrective force application areas with different thicknesses (Fig. 4). Corrective forces are distributed to the surface of the block structure, thereby achieving asymmetric stiffness in both concave and convex sides.
In particular, in the middle of the upper and lower corrective force application areas of the concave side of the brace, an outward block-structure lump with a height of 6 mm is designed, with the gap distance between the lumps. When the block structures are pressed against each other, a supporting force is formed to prevent further lateral bending of the spine. The gaps between the massive structures provide protection for the wearer during exercise for a certain range of muscle movement, aiding in the faster corrective recovery of the spine. Corrective pressure to the lump area of the torso causes the soft tissue to shift from the convex side (pressure zone) to the concave side (relief space) and the force is transmitted to the vertebral body to correct scoliosis.
The design of the brace on the convex side of the spine (mainly the lumbar and thoracic vertebrae) is a block structure adapted to the geometric epidermis of the patient torso and the applied pressure is a vector force with direction and magnitude, which applies a traction force on the sagittal plane, as well as a derotation force on the transverse plane. The height of the block structure and the tightness of the ventral strap determine the correct pressure and relief space of the orthosis.
The advantages of the block structure are twofold: first, the corrective force can be accurately transmitted to the corrective area of the patient torso. Second, the block structure is attached to the base with different thicknesses on the concave and convex sides, which can ensure that the rigidity of the two sides of the orthosis is different.
2.2 Finite Element Analysis.
A 3D FE model of the brace was implemented to investigate the static stiffness of the concave and convex sides of the brace. The 3D model of the brace was a two-dimensional–3D-solid map meshed by 90,688 eight-node hexahedral elements (C3D8R) with a characteristic length of 2 mm in Hypermesh2019.1 (Altair, Inc., France) (Fig. 5(a)). All meshing and simulations were run in abaqus 6.14 (DASSAULT, Inc., France) software. According to the pain thresholds, the applied range of force was 0–100 N, and the applied area was 64–225 cm2 [20,21]. The maximum pain threshold of human body is 70 N . The patient in the present research was a one-year-old female child. According to rule of thumb, the pain thresholds that can be endured by the elderly and children are 30% lower than those by the corresponding adults, respectively. In addition, pain thresholds are the results of subjective evaluations and may vary from person to person. Before finite element analysis (FEA), pain threshold tests on the patients in this paper were performed by doctors. Application of over 50 N on the convex side and concave sides of the patient produced discomfort. Therefore, in this study, a total force of 50 N was applied on the orthopedic force application area, totaling 92.96 cm2 on both sides (i.e., 33.12 cm2 for the convex side and 59.84 cm2 for the concave side) (Fig. 5(b)).
The predicted distribution of contact pressure was compared with published data of experimental measurements obtained under the same boundary conditions as validation.
2.3 Fabrication Using Three-Dimensional Printing Technology.
The brace was then fabricated using 3D printing technology and applied in the clinic. The mechanical properties could be adjusted with various brace characteristic parameters, such as the unit structure and cross-sectional size of the struts. The filling ratio of the brace in this study was 100%, which provided strong support and relatively flexible stiffness to the paravertebral muscles and spine. The thermoplastic polyurethane (TPU), which was used as the brace material in this study, is a homogeneous and isotropic elastic material with a Young's modulus of 27 MPa and a Poisson's ratio of 0.3 . The density of TPU is low, the weight and thickness can be adjusted within the ideal range and has a certain flexibility, compared with photosensitive resin and other 3D printing materials. Fused deposition modeling (FDM) is a 3D-printing method that can provide the desired strength and material deposition rate for custom brace applications. Proper selection of material for manufacturing of brace leads to greater increase in popularity in patients during daily life, providing a better patient-specific compatible brace (Fig. 6).
2.4 Experimental Verification: Brace Wear and Corrective Force Measurement.
To evaluate the asymmetric stiffness characteristics of the brace and aid in brace wear adherence, the brace was equipped with a thin-film pressure sensor (Fig. 7). Eight block-structure lumps existed on the convex and concave sides of the brace, respectively. Each block-structure lump was affixed with a thin-film pressure sensor and eight recorded values on both sides were obtained each time. The average correction pressure on the concave and convex sides was obtained by averaging the eight sensor values, respectively. Note that the recorded pressure value was only a relative value and did not represent the magnitude of the actual force. The patient was tested in a natural sitting posture, while wearing the brace and connecting sensors. The microcontroller tested and recorded data every second. A total of 26 groups of data were recorded, and the results are shown in Fig. 8.
3.1 Static Stiffness Characteristic of Brace.
The simulation results showed that the stiffness of the concave and convex sides were 145.88 N/mm and 35.96 N/mm, respectively (Figs. 9(a) and 9(b)). The difference between the static stiffness characteristic between the concave and convex sides of the 3D printed patient-specific asymmetric stiffness brace were considerable.
In Fig. 8, the abscissa represents the test time, and the ordinate represents the relative pressure value. Corrective force measurement results showed that the total average corrective force relative value in the convex side was 33.53 N, while in the concave side the force was 3.27 N. Generally, the corrective forces on both sides of the brace are equal. However, the asymmetric stiffness of the concave and convex sides of the brace produced an unequal corrective force due to the asymmetric configuration. Because the stiffness of the right block structure is lower, more corrective force can be accurately transmitted to the convex side. In contrast, the stiffness of the concave side was higher, and the corrective force was mostly shielded, so the measured corrective force was less. That points to the fact that the corrective pressure distribution of the manufactured brace was similar to the simulation results of the design.
3.2 Finite Element Analysis Results.
Figure 5(c) stress cloud results showed that the maximum pressure was 2090 kPa at the area of where the straps come into contact with the hole. The maximum stress was the force between the straps and the hole, rather than the corrective force between braces and the body of the patient. Pressure between brace and body surface was 162.5 kPa in the FE model, the predicted pressure on the trunk of patient wearing the brace was consistent with the literature results [20,21,24]. Correction forces were guaranteed to range below the human pain threshold, which meets the design requirements. Moreover, the comparisons of FEA versus corrective force measurements and FEA versus existed results showed an agreement between the predicted and measured correct force when the patient was in-brace in the present research, pointing to the reliability of the FE model developed in this study.
3.3 Clinical Results.
Radiography remains the standard in curve progression documentation. The scoliosis Cobb angles before the treatment and at a six-month follow-up were measured from X-ray images. The clinical practice results show that the Cobb angle was reduced from 49° to 42.4° after 6 months of day-time correctional wearing of our asymmetric stiffness brace (Fig. 10).
According to the parents of the patient, the new asymmetric stiffness brace was more comfortable to wear, in comparison with a conventional one, making it more suitable for a child. No pressure sores were observed during the six-month correction. The body of the patient was free to move within a certain range, while wearing the brace, and was able to perform spinal exercises (for example, yoga), which is conducive to the correction of scoliosis and can prevent the progression of it. The new brace produced by 3D printing in combination with 3D scanning data can realize personalized customization at different stages of orthopedics, ensuring more targeted and accurate correction, compared with conventional braces.
This paper presented a new method for a 3D printed brace design that considers the asymmetric stiffness properties of the concave and convex paravertebral muscles. This proposed brace design method can be potentially useful in patient treatment acceptance. The study presented a novel therapy tool for EOS patients. This 3D custom-made brace features an improved brace design with a unique shift/push combination of forces that guide the spine into the corrected position.
Paravertebral muscles are thought to be involved in the onset and development of scoliosis and the biomechanical characteristics of paravertebral muscles in EOS were asymmetric between the concave and convex sides. Although various braces have been developed and applied clinically, there are few studies that focus on asymmetric complementary balance of the biomechanical properties, following the fact that the asymmetric development of paravertebral muscles leads to an imbalance in muscle strength, which in turn contributes to the development of scoliosis.
Early onset scoliosis is a scoliosis disease, which is common and complicated in diagnosis and treatment. Due to the young age of onset, the patient is still in a critical period of growth and development. Control and treatment of scoliosis during the spine and thoracic cavity development period is the most important and highly difficult aspect of corrective treatment. To the knowledge of the authors, this is the first FEA study to be published that considers the asymmetric stiffness properties of the concave and convex paravertebral muscles in EOS. Existing FE models in the literature mostly target adolescent idiopathic scoliosis. The patient in this paper was so young that surgical treatment would be ineffective or not as effective as conservative orthopedic treatment. To prevent the worsening of the scoliosis, while waiting for surgery, the low-cost customized brace with asymmetric stiffness was designed for this case, with the consent of the parents of the patient.
High pressure on the human body will not only produce discomfort but also cause pressure ulcers. To guarantee that the local pressure was within the tolerance range, a reasonable stress area needed to be selected. Using the same pressure thresholds for all patients remains a limitation because each person has a different tolerance. Pressure threshold data used for this study were collected from rule of thumb and clinical tests. However, even if pressure thresholds were respected, patients still felt discomfort. Despite this, pressure thresholds can still be used as a guide for brace design. The method of dividing the orthopedic force application area proposed in this paper, on the one hand, provided a quantitative orthopedic division method based on Cobb angles, bypassing the need to divide the force application based on experience. On the other hand, the quantitative division of the orthopedic force application area also provided a basis for orthopedic force falling into the range of human pain threshold.
Cobetto measured the pressure between brace and patient body surface and gave a tolerance threshold of 85–379 kPa. The torso was divided in nine anatomical regions, for which a corresponding specific threshold was matched . However, to the knowledge of the authors, no published studies exist that describe the optimal pressure distribution and maximum pressure that can be applied by braces regarding the comfort of a patient. Studies defining pressure pain thresholds for different anatomical regions indicate that body regions are not equally sensitive . However, these data do not consider scoliosis patient characteristics and brace design. The applied range of force was 0–100 N, and the applied area was 64–225 cm2. The applied force pressure between brace and patient body surface in the current research was 162.5 kPa, which was in agreement to the results of these studies.
Following investigation of the relationship between esthetics and brace design, Law found that codesigning with patients on the esthetic aspects of the surface design of the brace can increase the level of user compliance and induce positive user perception. By including esthetic options in the visual design, the physical discomfort created by the elements of the brace is psychologically remedied . Therefore, esthetic options need to be considered in the design process of braces . Patient acceptance for the spinal orthoses is always a concern as it could greatly affect the patient treatment compliance and outcome. In the current design of the brace, poor brace compliance as a risk factor for curve progression remains the most critical problem . Material property, configuration, esthetic, and comfort design of the brace are also related with brace compliance . Articles exist that point to the fact that brace wear has a psychological and social burden, which should be factored into decision-making when the treated patient is likely to fail brace treatment [29–31]. Compliance with brace wear may still be supported by colors and pictures on the brace that are preferred by children . Psychological changes induced by the brace wearing, pain and conflicts in the school environment, influence poor brace compliance, low value of in-brace correction etc. Poor brace compliance is the most mentioned issue in all reported studies . The designed brace in this paper can be color and pattern customized according to the user's preferences. The outline pattern of the heat dissipation hole can also be selected by the user, according to personal preference. The brace compliance can be modified via mapping of patient data and the brace design is customizable.
The following limitations are present in this study. The brace was evaluated without integrating the brace FE model with the human FE model. The threshold of pain might vary with age, and more simulation parameter values need to be detailed. Though the brace simulation results for the static stiffness of the brace, no correlation was detected to the deformity correction. To conclude about the effectiveness and reliability, it would be essential to include more patients in the study and compare the results of this brace correction with the results obtained from conventional braces.
Future research will focus on the improving of the FE model and on the optimization of the design method. Additionally, considerable effort will be dedicated to completely develop the totally automated design and simulation process and relative computer-aided design software for customized braces. In future application, torso CAD data and biomechanical data from the patient, such as the stiffness of paravertebral muscles, will be required to generate the FEA simulation model and the initial brace, based on which the final printed customized brace can be obtained through an automated optimization procedure. Future work will focus on addressing these problems.
A novel low-cost 3D printed brace design method for early onset scoliosis was proposed and initially applied in the clinic. The clinical application of the day-time brace demonstrated that safe correction was provided with the advantages of low-cost, easy manipulation and being experience-independent. The patient-specific 3D printed brace has also a number of advantages over the conventional brace: (a) It has a lower profile so it is more discreet under clothing. (b) A variety of colorful materials are available, such as customized cartoon figures and cutouts that are more likely for a child to wear. (c) It is lightweight and easy to manufacture.
This work was supported by National Key Research and Development Program (No. 2018YFE0207900), ShaanXi Province Key Research and Development Program (No. 2021GXLH-Z-034), and Science and Technology Project of Beilin District (No. GX2125).