Keywords
3D printing - additive manufacturing - patient-specific anatomical brace - distal
radius fracture - scaphoid fracture
Conservative treatment of relatively simple distal radius fractures may provide functional
outcomes that are comparable to operative fixation.[1]
[2]
[3]
[4] However, plaster or synthetic casts molded onto a fractured wrist offer poor support.[5]
[6] A study using computed tomography scan has demonstrated major gap spaces between
the inner surface of the cast and the surface of the wrist.[7] Other factors, such as the position and length of joint immobilization, and the
type of fracture (involving or not involving the elbow or thumb), play a disputed
role.
Although anatomical fracture reduction is generally believed to improve outcomes,[8] recent studies challenge this for distal radius[9]
[10]
[11] and scaphoid fractures.[12]
[13]
[14] Surgical interventions are associated with an increased risk of surgery-related
complications, and there is a risk of fixation failure in osteoporotic bone.[15] Therefore, noninvasiveness, avoidance of complications, and patient comfort are
important considerations in fracture management. Developing a more effective solution
to maintain fracture reduction without the need for surgery would be a significant
medical advancement.
To date, casting is still the gold standard treatment of stable distal radius and
scaphoid fractures. However, despite advances in materials, casts continue to be cumbersome
and poorly tolerated by the patients, limiting their ability to dress normally and
perform daily activities.[16] This discomfort is particularly pronounced in young patients. In fact, many young
adults with scaphoid fractures remove their plaster before the recommended period
of immobilization has ended.[17] To overcome these limitations, three-dimensional (3D)-printed patient-specific anatomical
braces (PSABs) have emerged as a potential solution. These PSABs are custom-made to
perfectly match the patient's limb anatomy, providing better fracture contention and
reducing the risk of fracture displacement and final malunion.[18] Recent advances in scanning and 3D printing have made it possible. Chen et al conducted
two pioneering studies on the application of a 3D-printed cast for distal radius fractures.[19]
[20] They showed that their novel casting technology heals the fracture effectively,
without casting complications. Their 3D-printed cast was patient specific, ventilated,
and lightweight, increasing patient comfort and satisfaction. While the computer-aided
design to obtain the STereoLithography (STL) format file for 3D printing can be time-consuming,
recent studies have reported improved scanning and design times.[21] Keller et al presented an in-hospital production of patient-specific 3D-printed
PSAB using a semiautomated modeling process and the use of photosensitive resin Digital
Light Processing (DLP) printing, achieving faster production time.[22] However, uploading the scans to an online platform owned by an industrial company
induces extra costs, and patient data protection must be top priority.
The purposes of the present study were to develop a model of wrist PSAB, adapted to
fracture care, with an automatic design system, to evaluate if this brace would be
well tolerated by healthy volunteers, and to determine its mechanical properties as
compared with conventional methods of wrist immobilization.
Materials and Methods
A consortium group of engineers from Idiap Research Institute and HES-SO University
of Applied Sciences, Fribourg, hand therapists and hand surgeons from the University
Hospital Bern, and from the MedTech Swibrace Ltd was set up with the aim of creating
a PSAB model that meets both technical and clinical requirements. By successive iterations,
the engineers proposed various models adapted to distal radius and scaphoid fractures
immobilization, sufficiently rigid to maintain fracture reduction, yet thin, lightweight,
and esthetically pleasing. Thanks to the lateral opening on the ulnar side of the
forearm and the semirigidity of the material, the PSAB model is able to adjust for
changes in mild posttraumatic swelling. Hand therapists and hand surgeons first tested
these prototypes for satisfaction, personal experience, and functionality during different
activities of daily living. Based on their feedback, improvements were made until
the final prototype was developed and tested in a preclinical study. Ethical approval
was obtained from the ethical committee in Bern, Switzerland (No. 2021-00112).
Creation of the “Adult-Rated Splint Evaluation Questionnaire”
Due to lack of validated orthosis satisfaction questionnaires suitable for our study
purposes,[23] we designed an original “Adult-Rated Splint Evaluation Questionnaire” (ARSEQ)[24] ([Supplementary Appendix I]). The ARSEQ included orthosis-related questions, function-specific questions based
on those used in the validated Patient-Rated Wrist Evaluation questionnaire,[25] and 3D technology-related questions. It was divided into five themes with several
subquestions each: (1) satisfaction and (2) personal experiences with the splint;
(3) specific (e.g., opening a door, using a mobile phone) and usual activities (e.g.,
doing household work, performing sports) in the splint; (4) personal attitude; and
(5) scanning procedure. Volunteers could indicate pressure marks and skin irritations
using pictures and photographs. The answers were rated on an 11-point Likert scale
(0–10). Since the number of questions varied between themes, points were expressed
as percentages. A mean satisfaction score of at least 70% was the threshold for considering
a brace acceptable for clinical use. For the preclinical study, its psychometric properties
were not yet validated.
Preclinical Study Procedures
For the preclinical study, a convenience sample of 10 healthy volunteers were recruited
by e-mail. Inclusion criteria were ≥18 years old, German language proficiency, and
no acute health problems affecting the hands.
Informed written consent was obtained before the procedure. The scanning was conducted
using a HandySCAN 300 (Creaform), capturing the entire forearm ([Fig. 1]). Volunteers chose which hand to scan and whether to test the brace model for radius
or scaphoid fractures, with the latter also including the metacarpal and the proximal
phalanx of the thumb. Subsequently, 3D printing of the splint was done using polyamide
PA12 material and Selective Laser Sintering (SLS) printing technology ([Table 1]). With this printing technology, no further postprocessing is necessary.
Fig. 1 Scanning procedure for a scaphoid splint. (A) Scanning the forearm with HandySCAN 300, reflecting self-adhesive dots are positioned
on the arm to increase scanning quality. (B) Virtual forearm model. (C) Patient-specific anatomical brace model ready to be printed.
Table 1
3D splint properties
|
Surface scanner
|
HandySCAN 300 from the company Creaform
|
|
3D design software
|
VXelements from the company Creaform
|
|
Printing material
|
Polyamide PA12, biocompatible
|
|
Printing technique
|
Selective laser sintering
|
|
Layer thickness
|
2.0 mm
|
|
Postprocessing
|
Automatic
|
|
Favorable mechanical properties of printed splints
|
Lightweight, smooth, semirigid, long-term stability
|
Abbreviation: 3D, three-dimensional.
The splint was worn directly on the skin for a planned duration of 72 hours without
interposing cotton. Volunteers were instructed to pursue their normal life including
work, activities of daily living, sports, and sleep. To document their activities,
they received a diary where they also noted whether they took off the splint occasionally
and why (e.g., while driving a car). Upon completing the splint wearing period, volunteers
filled out the ARSEQ.
The outcome measurements (ARSEQ and the activity diary) in the preclinical study were
analyzed, with results reported as percentages for each theme and frequencies for
each activity noted in the diary.
Mechanical Testing of the PSAB
A testing bench was constructed, allowing mechanical testing of the PSAB on an articulated,
3D-printed wrist and forearm mannequin ([Fig. 2A]). Measurements were performed using a hydraulic Instron tensile testing machine
with a 1 kN force cell, measuring both wrist flexion and extension. Brace stiffness,
expressed in Nm/degree, was determined as resulting bending moment for a given rotation
angle relative to the wrist joint. [Fig. 2B] shows typical stiffness curves for different brace fastening systems. Since the
stiffness curve behavior is nonlinear, a reference bending moment of 3.4 Nm was used
for stiffness determination, corresponding approximately to a 4 kg mass that the patient
would hold in the hand.
Fig. 2 (A) Test bench used to measure the brace stiffness in flexion and extension (forced
wrist extension illustrated). (B) Stiffness curves for forced extension, representing the bending moment as function
of the angular rotation: 3D splint with cable ties (red), lace (yellow), circular
SoftCAST, split SoftCAST with bandage, and prefabricated ManuLoc splint. Reference
bending moment for stiffness value determination (dashed red horizontal line).
Three novel tests for mechanical stiffness were performed. The first compared the
stiffness of braces fabricated with different materials and different 3D printing
technologies: brace 1, fabricated using SLS technology from polyamide PA12 (Materialise,
Belgium) and aged 2 years at testing time; brace 2, fabricated using SLS technology
from PA12 (Chromos group, Switzerland) and aged 1 month at testing time; and brace
3, fabricated using DLP from resin reactive urethane photopolymers Ultracur3D ST45
(Production ToGo GmbH, Germany) and aged 1 month at testing time. The second test,
employing a PA12 brace printed by Chromos, examined the influence of different attachment
methods on brace stiffness: three cable ties, three hook and loop strips, and one
lace. The third test compared the stiffness of the latter brace to other types of
immobilizations: plaster cast, with or without cutting, and a commercial over-the-counter
wrist brace (ManuLoc, Bauerfeind, Lena, Germany—[Fig. 2B]).
Automatic Design of the PSAB
In parallel, a mathematical algorithm was developed, facilitating the adaptation of
a standard model of the brace to the patient's anatomy. The approach used nonrigid
registration to deform a template limb mesh.[26] The template carried landmarks and information used in an automated workflow, allowing
for automatic design of the PSAB in a few minutes, based on the scanned morphology
of the patient's anatomy and on the selected existing standard brace model (radius
or scaphoid fracture model).
Results
Various models of PSAB were tested ([Figs. 3] and [4]). Prototype 1, with large open areas, was poorly tolerated, with an average satisfaction
rate reaching 56%, due to pressure marks that caused pain and numbness ([Fig. 3A–D]). Prototype 2 was too tight, also causing skin irritations and pressure marks, with
an average satisfaction rate of 62% ([Fig. 3E–H]). Prototype 3 fared better, receiving an overall satisfaction score of 68% ([Fig. 4A, B]).
Fig. 3 First two prototypes tested by hand therapists from the University Hospital in Bern
before testing on healthy volunteers. (A) Dorsal and (B) palmar views of the first prototype. (C) Pressure marks and (D) numbness of the thumb after a few hours of splint wearing time. (E) Dorsal and (F) ulnodorsal view of the second prototype. (G, H) Pressure marks after 5 hours of splint wearing time.
Fig. 4 Prototypes tested by healthy volunteers. (A) PSAB for radius fractures. (B) Closing system with an elastic lace. (C) Volunteers' feedback of the ARSEQ. Final PSAB model from (D) dorsal and (E) palmar views with hook-and-loop straps. (F) Volunteers' final ARSEQ scores. ARSEQ, Adult-Rated Splint Evaluation Questionnaire;
PSAB, patient-specific anatomical brace.
The mean age of the 10 healthy volunteers (6 men, 4 women, all right-handed) was 46 ± 18
years. The group included four blue-collar workers, three white-collar workers, two
retirees, and a student. Half of them wanted to test the PSAB on their dominant hand.
Five volunteers tested the radius fracture model, and the other five tested the scaphoid
fracture model.
The scaphoid fracture model scored worse than the radius, due to longer scanning time,
and because having the thumb immobilized negatively affected function ([Fig. 4C]). The main drawback was the closing system of the PSAB, which was too cumbersome.
Prototype 4, featuring a more user-friendly hook-and-loop strap closing system ([Fig. 4D, E]), achieved an overall satisfaction score of 79% (70% in the scaphoid and 87% in
the radius groups), exceeding the previously set cutoff of at least 70% for clinical
testing ([Fig. 4F]).
Volunteers reported of being able to wear the PSAB for all self-care activities including
taking a shower, eating, and sleeping. It was worn during work and housekeeping tasks,
such as typing on a computer, driving a forklift, or cooking ([Fig. 5]).
Fig. 5 Example of a cooking activity of a volunteer wearing the prototype 3 for radius fracture
on the nondominant hand.
Mechanical Testing (Prototype 4)
The first test demonstrated that the brace's mechanical behavior (torque vs. angular
displacement relation) was more linear and smoother for extension movement across
all materials and printing techniques. The aging effect of PA12 on the material's
mechanical properties was not significant, with the 2-year-old PA12 material being
stiffer in flexion but more flexible in extension. The DLP brace had the highest stiffness
in both flexion and extension ([Table 2]).
Table 2
Comparison of the stiffness of PSAB with different materials
|
Stiffness measured with the secant between 0 and 3.4 Nm
|
Stiffness (Nm/deg)
|
|
Extension
|
Flexion
|
|
PA12, aging 2 y
|
0.79 ± 0.02
|
0.69 ± 0.008
|
|
PA12, aging 1 mo
|
0.80 ± 0.01
|
0.64 ± 0.009
|
|
Ultracur3D ST45, aging 1 mo
|
0.85 ± 0.01
|
0.77 ± 0.01
|
Abbreviation: PSAB, patient-specific anatomical braces.
The second test demonstrated that, with each fixation system, the brace had larger
stiffness and more linear torque–angular displacement for the extension movement.
In flexion, initial stiffness was smaller and increased with increasing angular deformations.
Cable ties were the fixation method offering the highest stiffness to the brace (1.7
times stiffer than hook-and-loop straps, [Table 3]), both for flexion and extension. The curve presented a quasi-isotropic stiffness.
Table 3
Comparison of the stiffness of PSAB with plaster cast used in orthopaedic practice
|
Stiffness measured with the secant between 0 and 3.4 Nm
|
Stiffness (Nm/deg)
|
|
Extension
|
Flexion
|
|
PSAB with three cable ties
|
1.16
|
1.01
|
|
PSAB with lace
|
0.99
|
0.74
|
|
Circular plaster cast
|
0.78
|
0.84
|
|
Cut plaster cast, closed with bandage
|
0.77
|
0.76
|
|
ManuLoc
|
n/a
|
0.42
|
Abbreviations: n/a, not available; PSAB, patient-specific anatomical brace.
The third test demonstrated that the stiffness of the PSAB was slightly higher than
the plaster cast and significantly better than the over-the-counter wrist brace ([Table 3]).
Characteristics (Prototype 4)
The final PSAB model weighed 60 to 90 g, made of PA12, had a 2-mm thickness, was fixed
by hook-and-loop straps (Velcro), and covered two-thirds of the forearm length. There
was a 1-mm clearance between the brace's inner and the skin. Stiffness ranged from
0.64 to 0.99 Nm/degree. The production cost per PSAB model was 430 CHF including the
printing and shipping of the orthosis. The average time between the scanning appointment
and the delivery date of the PSAB was 8 ± 2 days. The rather long production time
was due to the fact that the PSABs are printed by a company that operates SLS technology
printers 24 hours a day, resulting in waiting times during production. Another reason
was due to the postal shipping between Belgium and Switzerland and the therewith related
delivery delay.
Discussion
Cast immobilization in fracture treatment offers poor stabilization and may be complicated
by vascular, cutaneous, and neurological problems.[27] These complications are frequently attributed to high cast stiffness, unbalanced
pressure caused by the cast, and impecunious ventilation. In contrast, 3D-printed
casts are customized to the patient's anatomy, entirely well ventilated, lightweight,
allow for radiological control, and have adjustable mechanical properties by changing
their material, shape, thickness, structure, and type of fixation.[18]
[28]
PSABs seem to be particularly suitable for treating undisplaced or minimally displaced
fractures. In such fractures, it is important to avoid secondary displacement which
occurs in a significant proportion of the cases. For the acute stabilization of displaced
distal radius fractures, where major swelling is present or anticipated, thermoplastic
splints, allowing good molding and individual adjustments, may still be superior.
PSAB technology relies on scanning the limbs of patients. We used a highly accurate
scanner, as the accuracy of cheaper scanning systems such as tablets with optical
sensors was not sufficient. During scanning, patients need to be immobile for a few
minutes in the specific position of immobilization. In the case of a displaced fracture,
it is theoretically possible to scan the healthy contralateral limb and to mirror
the morphological data (reversed symmetry). By doing so, one assumes that there is
symmetrical morphology, which is not totally accurate. Janzing et al,[29] for example, found an average left–right wrist circumferential difference of 3 mm
(range 0–20 mm) in 100 healthy volunteers. Another option, to be considered in the
future, is to obtain selected anatomical data of the traumatized wrist and to base
the brace design on a database of wrist scans.
For fracture treatment, a further limitation of PSAB treatment is the delay of obtaining
the brace after the fracture due to design and additive manufacturing time. The newly
created mathematical algorithm in this study will allow the development of a web-based
software with immediate semiautomatic design, but the production unit cannot be installed
in a hospital, at least for PSABs of acceptable quality obtained by laser sintering
techniques. Fused deposition modeling printers can be installed in hospitals, but
the braces obtained by filament fuse are of poor quality and not adapted to fracture
care. DLP printing could be a possible compromise.[22]
In this study, the original questionnaire ARSEQ was developed to evaluate wrist cast
immobilization, as validated quality assessment tools are lacking.[21] The authors are aware of the potential risk of measurement error introduced by using
a self-designed questionnaire that is not validated yet. Further investigations are
necessary to test the psychometric properties of the ARSEQ.
Volunteers wore the brace during unrestricted activities including sports, which does
not correspond to the initial activity limitations a patient would have with a fractured
distal radius or scaphoid. Two other studies provide clinical results from healthy
volunteers wearing wrist braces (not including the thumb). In the study by Graham
et al,[30] 12 healthy individuals tested a 3D splint for 2 hours. With 50.8/100 points, their
satisfaction with the splint was lower than in this sample (79/100 points) ([Fig. 4F]). In the study by Janzing et al,[29] 10 healthy persons tested a 3D splint for 7 days. Their comfort in the splint was
good (80/100 points), which is comparable to the personal experiences (68–90/100)
made by our volunteers wearing the PSAB designed for radius fractures ([Fig. 4]). Our volunteers reported being more restricted in activities of daily living than
those by Janzing et al,[29] who indicated no activity restrictions. The difference might result from the variation
in age, our sample being on average younger (mean age 46 years) than in Janzing et
al (mean age 58 years).
The original wrist PSAB models were developed for distal radius and scaphoid fractures,
with the latter immobilizing the metacarpal and proximal phalanxes of the thumb as
well. Both models are comfortable, elegant, and lightweight. The brace can be worn
without any padding, as only minor skin irritations were reported by the volunteers
on anatomical crucial points during specific movements (e.g., on the ulnar head when
pronating the forearm; in the first commissure when grasping a small object in the
scaphoid splint). The absence of padding enhances ventilation of the skin, prevents
unpleasant smell from the splint, and allows for water contact, which is an advantage
over conventional casts. Furthermore, the smooth surface and thinness of the 3D-printed
material provide high wearing comfort, not only for the skin but also for the clothes.
Despite its thinness, the brace offers better rigidity for fracture immobilization
than conventional splints. It is hoped that this brace will enable better fracture
stabilization and allow patients to continue with most of their daily activities despite
their injury. The final PSAB model is currently being tested on a series of patients
presenting stable distal radius or scaphoid fractures.
Conclusion
This collaborative research led to the development of a lightweight yet elegant PSAB,
adapted to fracture care. The brace provides more rigid wrist immobilization than
over-the-shelf splints or conventional casts. It is hoped that as the brace is more
anatomical, without gap space under the brace, there will be less fracture secondary
displacements, but this remains to be clinically demonstrated. PSABs represent the
future of orthopaedic immobilization, not only for fractures but also in other conditions,
such as degenerative or inflammatory osteoarticular affections, tendon diseases, and
neurological conditions, among others.
